Comprehensive discovery of noncoding RNAs in acute myeloid leukemia cell transcriptomes.

PubMed

Zhang, Jin; Griffith, Malachi; Miller, Christopher A; Griffith, Obi L; Spencer, David H; Walker, Jason R; Magrini, Vincent; McGrath, Sean D; Ly, Amy; Helton, Nichole M; Trissal, Maria; Link, Daniel C; Dang, Ha X; Larson, David E; Kulkarni, Shashikant; Cordes, Matthew G; Fronick, Catrina C; Fulton, Robert S; Klco, Jeffery M; Mardis, Elaine R; Ley, Timothy J; Wilson, Richard K; Maher, Christopher A

2017-11-01

To detect diverse and novel RNA species comprehensively, we compared deep small RNA and RNA sequencing (RNA-seq) methods applied to a primary acute myeloid leukemia (AML) sample. We were able to discover previously unannotated small RNAs using deep sequencing of a library method using broader insert size selection. We analyzed the long noncoding RNA (lncRNA) landscape in AML by comparing deep sequencing from multiple RNA-seq library construction methods for the sample that we studied and then integrating RNA-seq data from 179 AML cases. This identified lncRNAs that are completely novel, differentially expressed, and associated with specific AML subtypes. Our study revealed the complexity of the noncoding RNA transcriptome through a combined strategy of strand-specific small RNA and total RNA-seq. This dataset will serve as an invaluable resource for future RNA-based analyses. Copyright © 2017 ISEH – Society for Hematology and Stem Cells. Published by Elsevier Inc. All rights reserved.

Large-scale Labeled Datasets to Fuel Earth Science Deep Learning Applications

NASA Astrophysics Data System (ADS)

Maskey, M.; Ramachandran, R.; Miller, J.

2017-12-01

Deep learning has revolutionized computer vision and natural language processing with various algorithms scaled using high-performance computing. However, generic large-scale labeled datasets such as the ImageNet are the fuel that drives the impressive accuracy of deep learning results. Large-scale labeled datasets already exist in domains such as medical science, but creating them in the Earth science domain is a challenge. While there are ways to apply deep learning using limited labeled datasets, there is a need in the Earth sciences for creating large-scale labeled datasets for benchmarking and scaling deep learning applications. At the NASA Marshall Space Flight Center, we are using deep learning for a variety of Earth science applications where we have encountered the need for large-scale labeled datasets. We will discuss our approaches for creating such datasets and why these datasets are just as valuable as deep learning algorithms. We will also describe successful usage of these large-scale labeled datasets with our deep learning based applications.

Deep learning methods for protein torsion angle prediction.

PubMed

Li, Haiou; Hou, Jie; Adhikari, Badri; Lyu, Qiang; Cheng, Jianlin

2017-09-18

Deep learning is one of the most powerful machine learning methods that has achieved the state-of-the-art performance in many domains. Since deep learning was introduced to the field of bioinformatics in 2012, it has achieved success in a number of areas such as protein residue-residue contact prediction, secondary structure prediction, and fold recognition. In this work, we developed deep learning methods to improve the prediction of torsion (dihedral) angles of proteins. We design four different deep learning architectures to predict protein torsion angles. The architectures including deep neural network (DNN) and deep restricted Boltzmann machine (DRBN), deep recurrent neural network (DRNN) and deep recurrent restricted Boltzmann machine (DReRBM) since the protein torsion angle prediction is a sequence related problem. In addition to existing protein features, two new features (predicted residue contact number and the error distribution of torsion angles extracted from sequence fragments) are used as input to each of the four deep learning architectures to predict phi and psi angles of protein backbone. The mean absolute error (MAE) of phi and psi angles predicted by DRNN, DReRBM, DRBM and DNN is about 20-21° and 29-30° on an independent dataset. The MAE of phi angle is comparable to the existing methods, but the MAE of psi angle is 29°, 2° lower than the existing methods. On the latest CASP12 targets, our methods also achieved the performance better than or comparable to a state-of-the art method. Our experiment demonstrates that deep learning is a valuable method for predicting protein torsion angles. The deep recurrent network architecture performs slightly better than deep feed-forward architecture, and the predicted residue contact number and the error distribution of torsion angles extracted from sequence fragments are useful features for improving prediction accuracy.

An efficient annotation and gene-expression derivation tool for Illumina Solexa datasets.

PubMed

Hosseini, Parsa; Tremblay, Arianne; Matthews, Benjamin F; Alkharouf, Nadim W

2010-07-02

The data produced by an Illumina flow cell with all eight lanes occupied, produces well over a terabyte worth of images with gigabytes of reads following sequence alignment. The ability to translate such reads into meaningful annotation is therefore of great concern and importance. Very easily, one can get flooded with such a great volume of textual, unannotated data irrespective of read quality or size. CASAVA, a optional analysis tool for Illumina sequencing experiments, enables the ability to understand INDEL detection, SNP information, and allele calling. To not only extract from such analysis, a measure of gene expression in the form of tag-counts, but furthermore to annotate such reads is therefore of significant value. We developed TASE (Tag counting and Analysis of Solexa Experiments), a rapid tag-counting and annotation software tool specifically designed for Illumina CASAVA sequencing datasets. Developed in Java and deployed using jTDS JDBC driver and a SQL Server backend, TASE provides an extremely fast means of calculating gene expression through tag-counts while annotating sequenced reads with the gene's presumed function, from any given CASAVA-build. Such a build is generated for both DNA and RNA sequencing. Analysis is broken into two distinct components: DNA sequence or read concatenation, followed by tag-counting and annotation. The end result produces output containing the homology-based functional annotation and respective gene expression measure signifying how many times sequenced reads were found within the genomic ranges of functional annotations. TASE is a powerful tool to facilitate the process of annotating a given Illumina Solexa sequencing dataset. Our results indicate that both homology-based annotation and tag-count analysis are achieved in very efficient times, providing researchers to delve deep in a given CASAVA-build and maximize information extraction from a sequencing dataset. TASE is specially designed to translate sequence data in a CASAVA-build into functional annotations while producing corresponding gene expression measurements. Achieving such analysis is executed in an ultrafast and highly efficient manner, whether the analysis be a single-read or paired-end sequencing experiment. TASE is a user-friendly and freely available application, allowing rapid analysis and annotation of any given Illumina Solexa sequencing dataset with ease.

A sparse autoencoder-based deep neural network for protein solvent accessibility and contact number prediction.

PubMed

Deng, Lei; Fan, Chao; Zeng, Zhiwen

2017-12-28

Direct prediction of the three-dimensional (3D) structures of proteins from one-dimensional (1D) sequences is a challenging problem. Significant structural characteristics such as solvent accessibility and contact number are essential for deriving restrains in modeling protein folding and protein 3D structure. Thus, accurately predicting these features is a critical step for 3D protein structure building. In this study, we present DeepSacon, a computational method that can effectively predict protein solvent accessibility and contact number by using a deep neural network, which is built based on stacked autoencoder and a dropout method. The results demonstrate that our proposed DeepSacon achieves a significant improvement in the prediction quality compared with the state-of-the-art methods. We obtain 0.70 three-state accuracy for solvent accessibility, 0.33 15-state accuracy and 0.74 Pearson Correlation Coefficient (PCC) for the contact number on the 5729 monomeric soluble globular protein dataset. We also evaluate the performance on the CASP11 benchmark dataset, DeepSacon achieves 0.68 three-state accuracy and 0.69 PCC for solvent accessibility and contact number, respectively. We have shown that DeepSacon can reliably predict solvent accessibility and contact number with stacked sparse autoencoder and a dropout approach.

FANSe2: a robust and cost-efficient alignment tool for quantitative next-generation sequencing applications.

PubMed

Xiao, Chuan-Le; Mai, Zhi-Biao; Lian, Xin-Lei; Zhong, Jia-Yong; Jin, Jing-Jie; He, Qing-Yu; Zhang, Gong

2014-01-01

Correct and bias-free interpretation of the deep sequencing data is inevitably dependent on the complete mapping of all mappable reads to the reference sequence, especially for quantitative RNA-seq applications. Seed-based algorithms are generally slow but robust, while Burrows-Wheeler Transform (BWT) based algorithms are fast but less robust. To have both advantages, we developed an algorithm FANSe2 with iterative mapping strategy based on the statistics of real-world sequencing error distribution to substantially accelerate the mapping without compromising the accuracy. Its sensitivity and accuracy are higher than the BWT-based algorithms in the tests using both prokaryotic and eukaryotic sequencing datasets. The gene identification results of FANSe2 is experimentally validated, while the previous algorithms have false positives and false negatives. FANSe2 showed remarkably better consistency to the microarray than most other algorithms in terms of gene expression quantifications. We implemented a scalable and almost maintenance-free parallelization method that can utilize the computational power of multiple office computers, a novel feature not present in any other mainstream algorithm. With three normal office computers, we demonstrated that FANSe2 mapped an RNA-seq dataset generated from an entire Illunima HiSeq 2000 flowcell (8 lanes, 608 M reads) to masked human genome within 4.1 hours with higher sensitivity than Bowtie/Bowtie2. FANSe2 thus provides robust accuracy, full indel sensitivity, fast speed, versatile compatibility and economical computational utilization, making it a useful and practical tool for deep sequencing applications. FANSe2 is freely available at http://bioinformatics.jnu.edu.cn/software/fanse2/.

Deep sequencing of the Camellia sinensis transcriptome revealed candidate genes for major metabolic pathways of tea-specific compounds

DOE Office of Scientific and Technical Information (OSTI.GOV)

Shi, CY; Yang, H; Wei, CL

Tea is one of the most popular non-alcoholic beverages worldwide. However, the tea plant, Camellia sinensis, is difficult to culture in vitro, to transform, and has a large genome, rendering little genomic information available. Recent advances in large-scale RNA sequencing (RNA-seq) provide a fast, cost-effective, and reliable approach to generate large expression datasets for functional genomic analysis, which is especially suitable for non-model species with un-sequenced genomes. Using high-throughput Illumina RNA-seq, the transcriptome from poly (A){sup +} RNA of C. sinensis was analyzed at an unprecedented depth (2.59 gigabase pairs). Approximate 34.5 million reads were obtained, trimmed, and assembled intomore » 127,094 unigenes, with an average length of 355 bp and an N50 of 506 bp, which consisted of 788 contig clusters and 126,306 singletons. This number of unigenes was 10-fold higher than existing C. sinensis sequences deposited in GenBank (as of August 2010). Sequence similarity analyses against six public databases (Uniprot, NR and COGs at NCBI, Pfam, InterPro and KEGG) found 55,088 unigenes that could be annotated with gene descriptions, conserved protein domains, or gene ontology terms. Some of the unigenes were assigned to putative metabolic pathways. Targeted searches using these annotations identified the majority of genes associated with several primary metabolic pathways and natural product pathways that are important to tea quality, such as flavonoid, theanine and caffeine biosynthesis pathways. Novel candidate genes of these secondary pathways were discovered. Comparisons with four previously prepared cDNA libraries revealed that this transcriptome dataset has both a high degree of consistency with previous EST data and an approximate 20 times increase in coverage. Thirteen unigenes related to theanine and flavonoid synthesis were validated. Their expression patterns in different organs of the tea plant were analyzed by RT-PCR and quantitative real time PCR (qRT-PCR). An extensive transcriptome dataset has been obtained from the deep sequencing of tea plant. The coverage of the transcriptome is comprehensive enough to discover all known genes of several major metabolic pathways. This transcriptome dataset can serve as an important public information platform for gene expression, genomics, and functional genomic studies in C. sinensis.« less

Deep sequencing of the Camellia sinensis transcriptome revealed candidate genes for major metabolic pathways of tea-specific compounds

PubMed Central

2011-01-01

Background Tea is one of the most popular non-alcoholic beverages worldwide. However, the tea plant, Camellia sinensis, is difficult to culture in vitro, to transform, and has a large genome, rendering little genomic information available. Recent advances in large-scale RNA sequencing (RNA-seq) provide a fast, cost-effective, and reliable approach to generate large expression datasets for functional genomic analysis, which is especially suitable for non-model species with un-sequenced genomes. Results Using high-throughput Illumina RNA-seq, the transcriptome from poly (A)+ RNA of C. sinensis was analyzed at an unprecedented depth (2.59 gigabase pairs). Approximate 34.5 million reads were obtained, trimmed, and assembled into 127,094 unigenes, with an average length of 355 bp and an N50 of 506 bp, which consisted of 788 contig clusters and 126,306 singletons. This number of unigenes was 10-fold higher than existing C. sinensis sequences deposited in GenBank (as of August 2010). Sequence similarity analyses against six public databases (Uniprot, NR and COGs at NCBI, Pfam, InterPro and KEGG) found 55,088 unigenes that could be annotated with gene descriptions, conserved protein domains, or gene ontology terms. Some of the unigenes were assigned to putative metabolic pathways. Targeted searches using these annotations identified the majority of genes associated with several primary metabolic pathways and natural product pathways that are important to tea quality, such as flavonoid, theanine and caffeine biosynthesis pathways. Novel candidate genes of these secondary pathways were discovered. Comparisons with four previously prepared cDNA libraries revealed that this transcriptome dataset has both a high degree of consistency with previous EST data and an approximate 20 times increase in coverage. Thirteen unigenes related to theanine and flavonoid synthesis were validated. Their expression patterns in different organs of the tea plant were analyzed by RT-PCR and quantitative real time PCR (qRT-PCR). Conclusions An extensive transcriptome dataset has been obtained from the deep sequencing of tea plant. The coverage of the transcriptome is comprehensive enough to discover all known genes of several major metabolic pathways. This transcriptome dataset can serve as an important public information platform for gene expression, genomics, and functional genomic studies in C. sinensis. PMID:21356090

GenomeGems: evaluation of genetic variability from deep sequencing data

PubMed Central

2012-01-01

Background Detection of disease-causing mutations using Deep Sequencing technologies possesses great challenges. In particular, organizing the great amount of sequences generated so that mutations, which might possibly be biologically relevant, are easily identified is a difficult task. Yet, for this assignment only limited automatic accessible tools exist. Findings We developed GenomeGems to gap this need by enabling the user to view and compare Single Nucleotide Polymorphisms (SNPs) from multiple datasets and to load the data onto the UCSC Genome Browser for an expanded and familiar visualization. As such, via automatic, clear and accessible presentation of processed Deep Sequencing data, our tool aims to facilitate ranking of genomic SNP calling. GenomeGems runs on a local Personal Computer (PC) and is freely available at http://www.tau.ac.il/~nshomron/GenomeGems. Conclusions GenomeGems enables researchers to identify potential disease-causing SNPs in an efficient manner. This enables rapid turnover of information and leads to further experimental SNP validation. The tool allows the user to compare and visualize SNPs from multiple experiments and to easily load SNP data onto the UCSC Genome browser for further detailed information. PMID:22748151

Metavisitor, a Suite of Galaxy Tools for Simple and Rapid Detection and Discovery of Viruses in Deep Sequence Data

PubMed Central

Vernick, Kenneth D.

2017-01-01

Metavisitor is a software package that allows biologists and clinicians without specialized bioinformatics expertise to detect and assemble viral genomes from deep sequence datasets. The package is composed of a set of modular bioinformatic tools and workflows that are implemented in the Galaxy framework. Using the graphical Galaxy workflow editor, users with minimal computational skills can use existing Metavisitor workflows or adapt them to suit specific needs by adding or modifying analysis modules. Metavisitor works with DNA, RNA or small RNA sequencing data over a range of read lengths and can use a combination of de novo and guided approaches to assemble genomes from sequencing reads. We show that the software has the potential for quick diagnosis as well as discovery of viruses from a vast array of organisms. Importantly, we provide here executable Metavisitor use cases, which increase the accessibility and transparency of the software, ultimately enabling biologists or clinicians to focus on biological or medical questions. PMID:28045932

SHARAKU: an algorithm for aligning and clustering read mapping profiles of deep sequencing in non-coding RNA processing.

PubMed

Tsuchiya, Mariko; Amano, Kojiro; Abe, Masaya; Seki, Misato; Hase, Sumitaka; Sato, Kengo; Sakakibara, Yasubumi

2016-06-15

Deep sequencing of the transcripts of regulatory non-coding RNA generates footprints of post-transcriptional processes. After obtaining sequence reads, the short reads are mapped to a reference genome, and specific mapping patterns can be detected called read mapping profiles, which are distinct from random non-functional degradation patterns. These patterns reflect the maturation processes that lead to the production of shorter RNA sequences. Recent next-generation sequencing studies have revealed not only the typical maturation process of miRNAs but also the various processing mechanisms of small RNAs derived from tRNAs and snoRNAs. We developed an algorithm termed SHARAKU to align two read mapping profiles of next-generation sequencing outputs for non-coding RNAs. In contrast with previous work, SHARAKU incorporates the primary and secondary sequence structures into an alignment of read mapping profiles to allow for the detection of common processing patterns. Using a benchmark simulated dataset, SHARAKU exhibited superior performance to previous methods for correctly clustering the read mapping profiles with respect to 5'-end processing and 3'-end processing from degradation patterns and in detecting similar processing patterns in deriving the shorter RNAs. Further, using experimental data of small RNA sequencing for the common marmoset brain, SHARAKU succeeded in identifying the significant clusters of read mapping profiles for similar processing patterns of small derived RNA families expressed in the brain. The source code of our program SHARAKU is available at http://www.dna.bio.keio.ac.jp/sharaku/, and the simulated dataset used in this work is available at the same link. Accession code: The sequence data from the whole RNA transcripts in the hippocampus of the left brain used in this work is available from the DNA DataBank of Japan (DDBJ) Sequence Read Archive (DRA) under the accession number DRA004502. yasu@bio.keio.ac.jp Supplementary data are available at Bioinformatics online. © The Author 2016. Published by Oxford University Press.

Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition.

PubMed

Ordóñez, Francisco Javier; Roggen, Daniel

2016-01-18

Human activity recognition (HAR) tasks have traditionally been solved using engineered features obtained by heuristic processes. Current research suggests that deep convolutional neural networks are suited to automate feature extraction from raw sensor inputs. However, human activities are made of complex sequences of motor movements, and capturing this temporal dynamics is fundamental for successful HAR. Based on the recent success of recurrent neural networks for time series domains, we propose a generic deep framework for activity recognition based on convolutional and LSTM recurrent units, which: (i) is suitable for multimodal wearable sensors; (ii) can perform sensor fusion naturally; (iii) does not require expert knowledge in designing features; and (iv) explicitly models the temporal dynamics of feature activations. We evaluate our framework on two datasets, one of which has been used in a public activity recognition challenge. Our results show that our framework outperforms competing deep non-recurrent networks on the challenge dataset by 4% on average; outperforming some of the previous reported results by up to 9%. Our results show that the framework can be applied to homogeneous sensor modalities, but can also fuse multimodal sensors to improve performance. We characterise key architectural hyperparameters' influence on performance to provide insights about their optimisation.

The sponge microbiome project.

PubMed

Moitinho-Silva, Lucas; Nielsen, Shaun; Amir, Amnon; Gonzalez, Antonio; Ackermann, Gail L; Cerrano, Carlo; Astudillo-Garcia, Carmen; Easson, Cole; Sipkema, Detmer; Liu, Fang; Steinert, Georg; Kotoulas, Giorgos; McCormack, Grace P; Feng, Guofang; Bell, James J; Vicente, Jan; Björk, Johannes R; Montoya, Jose M; Olson, Julie B; Reveillaud, Julie; Steindler, Laura; Pineda, Mari-Carmen; Marra, Maria V; Ilan, Micha; Taylor, Michael W; Polymenakou, Paraskevi; Erwin, Patrick M; Schupp, Peter J; Simister, Rachel L; Knight, Rob; Thacker, Robert W; Costa, Rodrigo; Hill, Russell T; Lopez-Legentil, Susanna; Dailianis, Thanos; Ravasi, Timothy; Hentschel, Ute; Li, Zhiyong; Webster, Nicole S; Thomas, Torsten

2017-10-01

Marine sponges (phylum Porifera) are a diverse, phylogenetically deep-branching clade known for forming intimate partnerships with complex communities of microorganisms. To date, 16S rRNA gene sequencing studies have largely utilised different extraction and amplification methodologies to target the microbial communities of a limited number of sponge species, severely limiting comparative analyses of sponge microbial diversity and structure. Here, we provide an extensive and standardised dataset that will facilitate sponge microbiome comparisons across large spatial, temporal, and environmental scales. Samples from marine sponges (n = 3569 specimens), seawater (n = 370), marine sediments (n = 65) and other environments (n = 29) were collected from different locations across the globe. This dataset incorporates at least 268 different sponge species, including several yet unidentified taxa. The V4 region of the 16S rRNA gene was amplified and sequenced from extracted DNA using standardised procedures. Raw sequences (total of 1.1 billion sequences) were processed and clustered with (i) a standard protocol using QIIME closed-reference picking resulting in 39 543 operational taxonomic units (OTU) at 97% sequence identity, (ii) a de novo clustering using Mothur resulting in 518 246 OTUs, and (iii) a new high-resolution Deblur protocol resulting in 83 908 unique bacterial sequences. Abundance tables, representative sequences, taxonomic classifications, and metadata are provided. This dataset represents a comprehensive resource of sponge-associated microbial communities based on 16S rRNA gene sequences that can be used to address overarching hypotheses regarding host-associated prokaryotes, including host specificity, convergent evolution, environmental drivers of microbiome structure, and the sponge-associated rare biosphere. © The Authors 2017. Published by Oxford University Press.

DSAP: deep-sequencing small RNA analysis pipeline.

PubMed

Huang, Po-Jung; Liu, Yi-Chung; Lee, Chi-Ching; Lin, Wei-Chen; Gan, Richie Ruei-Chi; Lyu, Ping-Chiang; Tang, Petrus

2010-07-01

DSAP is an automated multiple-task web service designed to provide a total solution to analyzing deep-sequencing small RNA datasets generated by next-generation sequencing technology. DSAP uses a tab-delimited file as an input format, which holds the unique sequence reads (tags) and their corresponding number of copies generated by the Solexa sequencing platform. The input data will go through four analysis steps in DSAP: (i) cleanup: removal of adaptors and poly-A/T/C/G/N nucleotides; (ii) clustering: grouping of cleaned sequence tags into unique sequence clusters; (iii) non-coding RNA (ncRNA) matching: sequence homology mapping against a transcribed sequence library from the ncRNA database Rfam (http://rfam.sanger.ac.uk/); and (iv) known miRNA matching: detection of known miRNAs in miRBase (http://www.mirbase.org/) based on sequence homology. The expression levels corresponding to matched ncRNAs and miRNAs are summarized in multi-color clickable bar charts linked to external databases. DSAP is also capable of displaying miRNA expression levels from different jobs using a log(2)-scaled color matrix. Furthermore, a cross-species comparative function is also provided to show the distribution of identified miRNAs in different species as deposited in miRBase. DSAP is available at http://dsap.cgu.edu.tw.

An efficient annotation and gene-expression derivation tool for Illumina Solexa datasets

PubMed Central

2010-01-01

Background The data produced by an Illumina flow cell with all eight lanes occupied, produces well over a terabyte worth of images with gigabytes of reads following sequence alignment. The ability to translate such reads into meaningful annotation is therefore of great concern and importance. Very easily, one can get flooded with such a great volume of textual, unannotated data irrespective of read quality or size. CASAVA, a optional analysis tool for Illumina sequencing experiments, enables the ability to understand INDEL detection, SNP information, and allele calling. To not only extract from such analysis, a measure of gene expression in the form of tag-counts, but furthermore to annotate such reads is therefore of significant value. Findings We developed TASE (Tag counting and Analysis of Solexa Experiments), a rapid tag-counting and annotation software tool specifically designed for Illumina CASAVA sequencing datasets. Developed in Java and deployed using jTDS JDBC driver and a SQL Server backend, TASE provides an extremely fast means of calculating gene expression through tag-counts while annotating sequenced reads with the gene's presumed function, from any given CASAVA-build. Such a build is generated for both DNA and RNA sequencing. Analysis is broken into two distinct components: DNA sequence or read concatenation, followed by tag-counting and annotation. The end result produces output containing the homology-based functional annotation and respective gene expression measure signifying how many times sequenced reads were found within the genomic ranges of functional annotations. Conclusions TASE is a powerful tool to facilitate the process of annotating a given Illumina Solexa sequencing dataset. Our results indicate that both homology-based annotation and tag-count analysis are achieved in very efficient times, providing researchers to delve deep in a given CASAVA-build and maximize information extraction from a sequencing dataset. TASE is specially designed to translate sequence data in a CASAVA-build into functional annotations while producing corresponding gene expression measurements. Achieving such analysis is executed in an ultrafast and highly efficient manner, whether the analysis be a single-read or paired-end sequencing experiment. TASE is a user-friendly and freely available application, allowing rapid analysis and annotation of any given Illumina Solexa sequencing dataset with ease. PMID:20598141

QQ-SNV: single nucleotide variant detection at low frequency by comparing the quality quantiles.

PubMed

Van der Borght, Koen; Thys, Kim; Wetzels, Yves; Clement, Lieven; Verbist, Bie; Reumers, Joke; van Vlijmen, Herman; Aerssens, Jeroen

2015-11-10

Next generation sequencing enables studying heterogeneous populations of viral infections. When the sequencing is done at high coverage depth ("deep sequencing"), low frequency variants can be detected. Here we present QQ-SNV (http://sourceforge.net/projects/qqsnv), a logistic regression classifier model developed for the Illumina sequencing platforms that uses the quantiles of the quality scores, to distinguish true single nucleotide variants from sequencing errors based on the estimated SNV probability. To train the model, we created a dataset of an in silico mixture of five HIV-1 plasmids. Testing of our method in comparison to the existing methods LoFreq, ShoRAH, and V-Phaser 2 was performed on two HIV and four HCV plasmid mixture datasets and one influenza H1N1 clinical dataset. For default application of QQ-SNV, variants were called using a SNV probability cutoff of 0.5 (QQ-SNV(D)). To improve the sensitivity we used a SNV probability cutoff of 0.0001 (QQ-SNV(HS)). To also increase specificity, SNVs called were overruled when their frequency was below the 80(th) percentile calculated on the distribution of error frequencies (QQ-SNV(HS-P80)). When comparing QQ-SNV versus the other methods on the plasmid mixture test sets, QQ-SNV(D) performed similarly to the existing approaches. QQ-SNV(HS) was more sensitive on all test sets but with more false positives. QQ-SNV(HS-P80) was found to be the most accurate method over all test sets by balancing sensitivity and specificity. When applied to a paired-end HCV sequencing study, with lowest spiked-in true frequency of 0.5%, QQ-SNV(HS-P80) revealed a sensitivity of 100% (vs. 40-60% for the existing methods) and a specificity of 100% (vs. 98.0-99.7% for the existing methods). In addition, QQ-SNV required the least overall computation time to process the test sets. Finally, when testing on a clinical sample, four putative true variants with frequency below 0.5% were consistently detected by QQ-SNV(HS-P80) from different generations of Illumina sequencers. We developed and successfully evaluated a novel method, called QQ-SNV, for highly efficient single nucleotide variant calling on Illumina deep sequencing virology data.

Evidence for a persistent microbial seed bank throughout the global ocean

PubMed Central

Gibbons, Sean M.; Caporaso, J. Gregory; Pirrung, Meg; Field, Dawn; Knight, Rob; Gilbert, Jack A.

2013-01-01

Do bacterial taxa demonstrate clear endemism, like macroorganisms, or can one site’s bacterial community recapture the total phylogenetic diversity of the world’s oceans? Here we compare a deep bacterial community characterization from one site in the English Channel (L4-DeepSeq) with 356 datasets from the International Census of Marine Microbes (ICoMM) taken from around the globe (ranging from marine pelagic and sediment samples to sponge-associated environments). At the L4-DeepSeq site, increasing sequencing depth uncovers greater phylogenetic overlap with the global ICoMM data. This site contained 31.7–66.2% of operational taxonomic units identified in a given ICoMM biome. Extrapolation of this overlap suggests that 1.93 × 1011 sequences from the L4 site would capture all ICoMM bacterial phylogenetic diversity. Current technology trends suggest this limit may be attainable within 3 y. These results strongly suggest the marine biosphere maintains a previously undetected, persistent microbial seed bank. PMID:23487761

Deep learning for pharmacovigilance: recurrent neural network architectures for labeling adverse drug reactions in Twitter posts.

PubMed

Cocos, Anne; Fiks, Alexander G; Masino, Aaron J

2017-07-01

Social media is an important pharmacovigilance data source for adverse drug reaction (ADR) identification. Human review of social media data is infeasible due to data quantity, thus natural language processing techniques are necessary. Social media includes informal vocabulary and irregular grammar, which challenge natural language processing methods. Our objective is to develop a scalable, deep-learning approach that exceeds state-of-the-art ADR detection performance in social media. We developed a recurrent neural network (RNN) model that labels words in an input sequence with ADR membership tags. The only input features are word-embedding vectors, which can be formed through task-independent pretraining or during ADR detection training. Our best-performing RNN model used pretrained word embeddings created from a large, non-domain-specific Twitter dataset. It achieved an approximate match F-measure of 0.755 for ADR identification on the dataset, compared to 0.631 for a baseline lexicon system and 0.65 for the state-of-the-art conditional random field model. Feature analysis indicated that semantic information in pretrained word embeddings boosted sensitivity and, combined with contextual awareness captured in the RNN, precision. Our model required no task-specific feature engineering, suggesting generalizability to additional sequence-labeling tasks. Learning curve analysis showed that our model reached optimal performance with fewer training examples than the other models. ADR detection performance in social media is significantly improved by using a contextually aware model and word embeddings formed from large, unlabeled datasets. The approach reduces manual data-labeling requirements and is scalable to large social media datasets. © The Author 2017. Published by Oxford University Press on behalf of the American Medical Informatics Association. All rights reserved. For Permissions, please email: journals.permissions@oup.com

A visual tracking method based on deep learning without online model updating

NASA Astrophysics Data System (ADS)

Tang, Cong; Wang, Yicheng; Feng, Yunsong; Zheng, Chao; Jin, Wei

2018-02-01

The paper proposes a visual tracking method based on deep learning without online model updating. In consideration of the advantages of deep learning in feature representation, deep model SSD (Single Shot Multibox Detector) is used as the object extractor in the tracking model. Simultaneously, the color histogram feature and HOG (Histogram of Oriented Gradient) feature are combined to select the tracking object. In the process of tracking, multi-scale object searching map is built to improve the detection performance of deep detection model and the tracking efficiency. In the experiment of eight respective tracking video sequences in the baseline dataset, compared with six state-of-the-art methods, the method in the paper has better robustness in the tracking challenging factors, such as deformation, scale variation, rotation variation, illumination variation, and background clutters, moreover, its general performance is better than other six tracking methods.

Rapid Fine Conformational Epitope Mapping Using Comprehensive Mutagenesis and Deep Sequencing*

PubMed Central

Kowalsky, Caitlin A.; Faber, Matthew S.; Nath, Aritro; Dann, Hailey E.; Kelly, Vince W.; Liu, Li; Shanker, Purva; Wagner, Ellen K.; Maynard, Jennifer A.; Chan, Christina; Whitehead, Timothy A.

2015-01-01

Knowledge of the fine location of neutralizing and non-neutralizing epitopes on human pathogens affords a better understanding of the structural basis of antibody efficacy, which will expedite rational design of vaccines, prophylactics, and therapeutics. However, full utilization of the wealth of information from single cell techniques and antibody repertoire sequencing awaits the development of a high throughput, inexpensive method to map the conformational epitopes for antibody-antigen interactions. Here we show such an approach that combines comprehensive mutagenesis, cell surface display, and DNA deep sequencing. We develop analytical equations to identify epitope positions and show the method effectiveness by mapping the fine epitope for different antibodies targeting TNF, pertussis toxin, and the cancer target TROP2. In all three cases, the experimentally determined conformational epitope was consistent with previous experimental datasets, confirming the reliability of the experimental pipeline. Once the comprehensive library is generated, fine conformational epitope maps can be prepared at a rate of four per day. PMID:26296891

Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition

PubMed Central

Ordóñez, Francisco Javier; Roggen, Daniel

2016-01-01

Human activity recognition (HAR) tasks have traditionally been solved using engineered features obtained by heuristic processes. Current research suggests that deep convolutional neural networks are suited to automate feature extraction from raw sensor inputs. However, human activities are made of complex sequences of motor movements, and capturing this temporal dynamics is fundamental for successful HAR. Based on the recent success of recurrent neural networks for time series domains, we propose a generic deep framework for activity recognition based on convolutional and LSTM recurrent units, which: (i) is suitable for multimodal wearable sensors; (ii) can perform sensor fusion naturally; (iii) does not require expert knowledge in designing features; and (iv) explicitly models the temporal dynamics of feature activations. We evaluate our framework on two datasets, one of which has been used in a public activity recognition challenge. Our results show that our framework outperforms competing deep non-recurrent networks on the challenge dataset by 4% on average; outperforming some of the previous reported results by up to 9%. Our results show that the framework can be applied to homogeneous sensor modalities, but can also fuse multimodal sensors to improve performance. We characterise key architectural hyperparameters’ influence on performance to provide insights about their optimisation. PMID:26797612

An introduction to deep learning on biological sequence data: examples and solutions.

PubMed

Jurtz, Vanessa Isabell; Johansen, Alexander Rosenberg; Nielsen, Morten; Almagro Armenteros, Jose Juan; Nielsen, Henrik; Sønderby, Casper Kaae; Winther, Ole; Sønderby, Søren Kaae

2017-11-15

Deep neural network architectures such as convolutional and long short-term memory networks have become increasingly popular as machine learning tools during the recent years. The availability of greater computational resources, more data, new algorithms for training deep models and easy to use libraries for implementation and training of neural networks are the drivers of this development. The use of deep learning has been especially successful in image recognition; and the development of tools, applications and code examples are in most cases centered within this field rather than within biology. Here, we aim to further the development of deep learning methods within biology by providing application examples and ready to apply and adapt code templates. Given such examples, we illustrate how architectures consisting of convolutional and long short-term memory neural networks can relatively easily be designed and trained to state-of-the-art performance on three biological sequence problems: prediction of subcellular localization, protein secondary structure and the binding of peptides to MHC Class II molecules. All implementations and datasets are available online to the scientific community at https://github.com/vanessajurtz/lasagne4bio. skaaesonderby@gmail.com. Supplementary data are available at Bioinformatics online. © The Author (2017). Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Enhancing Hi-C data resolution with deep convolutional neural network HiCPlus.

PubMed

Zhang, Yan; An, Lin; Xu, Jie; Zhang, Bo; Zheng, W Jim; Hu, Ming; Tang, Jijun; Yue, Feng

2018-02-21

Although Hi-C technology is one of the most popular tools for studying 3D genome organization, due to sequencing cost, the resolution of most Hi-C datasets are coarse and cannot be used to link distal regulatory elements to their target genes. Here we develop HiCPlus, a computational approach based on deep convolutional neural network, to infer high-resolution Hi-C interaction matrices from low-resolution Hi-C data. We demonstrate that HiCPlus can impute interaction matrices highly similar to the original ones, while only using 1/16 of the original sequencing reads. We show that the models learned from one cell type can be applied to make predictions in other cell or tissue types. Our work not only provides a computational framework to enhance Hi-C data resolution but also reveals features underlying the formation of 3D chromatin interactions.

Detection of microRNAs in color space.

PubMed

Marco, Antonio; Griffiths-Jones, Sam

2012-02-01

Deep sequencing provides inexpensive opportunities to characterize the transcriptional diversity of known genomes. The AB SOLiD technology generates millions of short sequencing reads in color-space; that is, the raw data is a sequence of colors, where each color represents 2 nt and each nucleotide is represented by two consecutive colors. This strategy is purported to have several advantages, including increased ability to distinguish sequencing errors from polymorphisms. Several programs have been developed to map short reads to genomes in color space. However, a number of previously unexplored technical issues arise when using SOLiD technology to characterize microRNAs. Here we explore these technical difficulties. First, since the sequenced reads are longer than the biological sequences, every read is expected to contain linker fragments. The color-calling error rate increases toward the 3(') end of the read such that recognizing the linker sequence for removal becomes problematic. Second, mapping in color space may lead to the loss of the first nucleotide of each read. We propose a sequential trimming and mapping approach to map small RNAs. Using our strategy, we reanalyze three published insect small RNA deep sequencing datasets and characterize 22 new microRNAs. A bash shell script to perform the sequential trimming and mapping procedure, called SeqTrimMap, is available at: http://www.mirbase.org/tools/seqtrimmap/ antonio.marco@manchester.ac.uk Supplementary data are available at Bioinformatics online.

Seismic evidence of Messinian salt in opposite margins of West Mediterranean

NASA Astrophysics Data System (ADS)

Mocnik, Arianna; Camerlenghi, Angelo; Del Ben, Anna; Geletti, Riccardo; Wardell, Nigel; Zgur, Fabrizio

2015-04-01

The post drift Messinian Salinity Crisis (MSC) affected the whole Mediterranean basin, with deposition of evaporitic sequences in the deep basins, in the lower continental slopes, and in several shallower marginal basins; usually, in the continental margins, the MSC originated noticeable erosional truncations that locally cause important hiatuses in the pre-Messinian sequences, covered by the Plio-Quaternary sediments. In this work we focus on the MSC seismic signature of two new seismic datasets acquired in 2010 (West Sardinia offshore) and in 2012 (within the Eurofleet project SALTFLU in the South Balearic continental margin and the northern Algero abyssal plain). The "Messinian trilogy" recognized in the West-Mediterranean abyssal plain, is characterized by different seismic facies: the Lower evaporite Unit (LU), the salt Mobile Unit (MU) and the Upper evaporite mainly gypsiferous Unit (UU). Both seismic datasets show the presence of the Messinian trilogy also if the LU is not always clearly interpretable due to the strong seismic signal absorption by the halite layers; the salt thickness of the MU is similar in both the basins as also the thickness and stratigraphy of the UU. The Upper Unit (UU) is made up of a well reflecting package of about 10 reflectors, partially deformed by salt tectonic and characterized by a thin transparent layer that we interpreted as salt sequence inner the shallower part of the UU. Below the stratified UU, the MU exhibits a transparent layer in the deep basin and also on the foot of the slope, where a negative reflector, related to the high interval velocity of salt, marks its base. The halokinetic processes are not homogeneously distributed in the region, forming a great number of diapirs on the foot of the slope (due to the pression of the slided sediments) and giant domes toward the deep basin (due to the higher thickness of the Plio-quaternary sediments). This distribution seems to be related to the amount of salt and of the sedimentary cover. During the MSC the margins of the West Mediterranean Sea seem to be involved in some tectonic events probably connected to reactivation of normal faults and to the fast variation of the water load related to sea level fluctuations. The absence of calibrating boreholes in the deep Mediterranean basins and the hard penetration of seismic energy below the evaporitic layers, represent a limit for the knowledge of the geological evolution of the basins; the interpretation of the presented datasets could be a contribution to the comprehension of the evaporitic deposition and early-stage salt deformation during the MSC in the Mediterranean sea.

Metagenome sequencing and 98 microbial genomes from Juan de Fuca Ridge flank subsurface fluids

NASA Astrophysics Data System (ADS)

Jungbluth, Sean P.; Amend, Jan P.; Rappé, Michael S.

2017-03-01

The global deep subsurface biosphere is one of the largest reservoirs for microbial life on our planet. This study takes advantage of new sampling technologies and couples them with improvements to DNA sequencing and associated informatics tools to reconstruct the genomes of uncultivated Bacteria and Archaea from fluids collected deep within the Juan de Fuca Ridge subseafloor. Here, we generated two metagenomes from borehole observatories located 311 meters apart and, using binning tools, retrieved 98 genomes from metagenomes (GFMs). Of the GFMs, 31 were estimated to be >90% complete, while an additional 17 were >70% complete. Phylogenomic analysis revealed 53 bacterial and 45 archaeal GFMs, of which nearly all were distantly related to known cultivated isolates. In the GFMs, abundant Bacteria included Chloroflexi, Nitrospirae, Acetothermia (OP1), EM3, Aminicenantes (OP8), Gammaproteobacteria, and Deltaproteobacteria, while abundant Archaea included Archaeoglobi, Bathyarchaeota (MCG), and Marine Benthic Group E (MBG-E). These data are the first GFMs reconstructed from the deep basaltic subseafloor biosphere, and provide a dataset available for further interrogation.

Metagenome sequencing and 98 microbial genomes from Juan de Fuca Ridge flank subsurface fluids.

PubMed

Jungbluth, Sean P; Amend, Jan P; Rappé, Michael S

2017-03-28

The global deep subsurface biosphere is one of the largest reservoirs for microbial life on our planet. This study takes advantage of new sampling technologies and couples them with improvements to DNA sequencing and associated informatics tools to reconstruct the genomes of uncultivated Bacteria and Archaea from fluids collected deep within the Juan de Fuca Ridge subseafloor. Here, we generated two metagenomes from borehole observatories located 311 meters apart and, using binning tools, retrieved 98 genomes from metagenomes (GFMs). Of the GFMs, 31 were estimated to be >90% complete, while an additional 17 were >70% complete. Phylogenomic analysis revealed 53 bacterial and 45 archaeal GFMs, of which nearly all were distantly related to known cultivated isolates. In the GFMs, abundant Bacteria included Chloroflexi, Nitrospirae, Acetothermia (OP1), EM3, Aminicenantes (OP8), Gammaproteobacteria, and Deltaproteobacteria, while abundant Archaea included Archaeoglobi, Bathyarchaeota (MCG), and Marine Benthic Group E (MBG-E). These data are the first GFMs reconstructed from the deep basaltic subseafloor biosphere, and provide a dataset available for further interrogation.

Metagenome sequencing and 98 microbial genomes from Juan de Fuca Ridge flank subsurface fluids

PubMed Central

Jungbluth, Sean P.; Amend, Jan P.; Rappé, Michael S.

2017-01-01

The global deep subsurface biosphere is one of the largest reservoirs for microbial life on our planet. This study takes advantage of new sampling technologies and couples them with improvements to DNA sequencing and associated informatics tools to reconstruct the genomes of uncultivated Bacteria and Archaea from fluids collected deep within the Juan de Fuca Ridge subseafloor. Here, we generated two metagenomes from borehole observatories located 311 meters apart and, using binning tools, retrieved 98 genomes from metagenomes (GFMs). Of the GFMs, 31 were estimated to be >90% complete, while an additional 17 were >70% complete. Phylogenomic analysis revealed 53 bacterial and 45 archaeal GFMs, of which nearly all were distantly related to known cultivated isolates. In the GFMs, abundant Bacteria included Chloroflexi, Nitrospirae, Acetothermia (OP1), EM3, Aminicenantes (OP8), Gammaproteobacteria, and Deltaproteobacteria, while abundant Archaea included Archaeoglobi, Bathyarchaeota (MCG), and Marine Benthic Group E (MBG-E). These data are the first GFMs reconstructed from the deep basaltic subseafloor biosphere, and provide a dataset available for further interrogation. PMID:28350381

Protein contact prediction by integrating deep multiple sequence alignments, coevolution and machine learning.

PubMed

Adhikari, Badri; Hou, Jie; Cheng, Jianlin

2018-03-01

In this study, we report the evaluation of the residue-residue contacts predicted by our three different methods in the CASP12 experiment, focusing on studying the impact of multiple sequence alignment, residue coevolution, and machine learning on contact prediction. The first method (MULTICOM-NOVEL) uses only traditional features (sequence profile, secondary structure, and solvent accessibility) with deep learning to predict contacts and serves as a baseline. The second method (MULTICOM-CONSTRUCT) uses our new alignment algorithm to generate deep multiple sequence alignment to derive coevolution-based features, which are integrated by a neural network method to predict contacts. The third method (MULTICOM-CLUSTER) is a consensus combination of the predictions of the first two methods. We evaluated our methods on 94 CASP12 domains. On a subset of 38 free-modeling domains, our methods achieved an average precision of up to 41.7% for top L/5 long-range contact predictions. The comparison of the three methods shows that the quality and effective depth of multiple sequence alignments, coevolution-based features, and machine learning integration of coevolution-based features and traditional features drive the quality of predicted protein contacts. On the full CASP12 dataset, the coevolution-based features alone can improve the average precision from 28.4% to 41.6%, and the machine learning integration of all the features further raises the precision to 56.3%, when top L/5 predicted long-range contacts are evaluated. And the correlation between the precision of contact prediction and the logarithm of the number of effective sequences in alignments is 0.66. © 2017 Wiley Periodicals, Inc.

Improving protein disorder prediction by deep bidirectional long short-term memory recurrent neural networks.

PubMed

Hanson, Jack; Yang, Yuedong; Paliwal, Kuldip; Zhou, Yaoqi

2017-03-01

Capturing long-range interactions between structural but not sequence neighbors of proteins is a long-standing challenging problem in bioinformatics. Recently, long short-term memory (LSTM) networks have significantly improved the accuracy of speech and image classification problems by remembering useful past information in long sequential events. Here, we have implemented deep bidirectional LSTM recurrent neural networks in the problem of protein intrinsic disorder prediction. The new method, named SPOT-Disorder, has steadily improved over a similar method using a traditional, window-based neural network (SPINE-D) in all datasets tested without separate training on short and long disordered regions. Independent tests on four other datasets including the datasets from critical assessment of structure prediction (CASP) techniques and >10 000 annotated proteins from MobiDB, confirmed SPOT-Disorder as one of the best methods in disorder prediction. Moreover, initial studies indicate that the method is more accurate in predicting functional sites in disordered regions. These results highlight the usefulness combining LSTM with deep bidirectional recurrent neural networks in capturing non-local, long-range interactions for bioinformatics applications. SPOT-disorder is available as a web server and as a standalone program at: http://sparks-lab.org/server/SPOT-disorder/index.php . j.hanson@griffith.edu.au or yuedong.yang@griffith.edu.au or yaoqi.zhou@griffith.edu.au. Supplementary data is available at Bioinformatics online. © The Author 2016. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com

DRREP: deep ridge regressed epitope predictor.

PubMed

Sher, Gene; Zhi, Degui; Zhang, Shaojie

2017-10-03

The ability to predict epitopes plays an enormous role in vaccine development in terms of our ability to zero in on where to do a more thorough in-vivo analysis of the protein in question. Though for the past decade there have been numerous advancements and improvements in epitope prediction, on average the best benchmark prediction accuracies are still only around 60%. New machine learning algorithms have arisen within the domain of deep learning, text mining, and convolutional networks. This paper presents a novel analytically trained and string kernel using deep neural network, which is tailored for continuous epitope prediction, called: Deep Ridge Regressed Epitope Predictor (DRREP). DRREP was tested on long protein sequences from the following datasets: SARS, Pellequer, HIV, AntiJen, and SEQ194. DRREP was compared to numerous state of the art epitope predictors, including the most recently published predictors called LBtope and DMNLBE. Using area under ROC curve (AUC), DRREP achieved a performance improvement over the best performing predictors on SARS (13.7%), HIV (8.9%), Pellequer (1.5%), and SEQ194 (3.1%), with its performance being matched only on the AntiJen dataset, by the LBtope predictor, where both DRREP and LBtope achieved an AUC of 0.702. DRREP is an analytically trained deep neural network, thus capable of learning in a single step through regression. By combining the features of deep learning, string kernels, and convolutional networks, the system is able to perform residue-by-residue prediction of continues epitopes with higher accuracy than the current state of the art predictors.

Extracting features from protein sequences to improve deep extreme learning machine for protein fold recognition.

PubMed

Ibrahim, Wisam; Abadeh, Mohammad Saniee

2017-05-21

Protein fold recognition is an important problem in bioinformatics to predict three-dimensional structure of a protein. One of the most challenging tasks in protein fold recognition problem is the extraction of efficient features from the amino-acid sequences to obtain better classifiers. In this paper, we have proposed six descriptors to extract features from protein sequences. These descriptors are applied in the first stage of a three-stage framework PCA-DELM-LDA to extract feature vectors from the amino-acid sequences. Principal Component Analysis PCA has been implemented to reduce the number of extracted features. The extracted feature vectors have been used with original features to improve the performance of the Deep Extreme Learning Machine DELM in the second stage. Four new features have been extracted from the second stage and used in the third stage by Linear Discriminant Analysis LDA to classify the instances into 27 folds. The proposed framework is implemented on the independent and combined feature sets in SCOP datasets. The experimental results show that extracted feature vectors in the first stage could improve the performance of DELM in extracting new useful features in second stage. Copyright © 2017 Elsevier Ltd. All rights reserved.

A Comprehensive Phylogenetic Analysis of the Scleractinia (Cnidaria, Anthozoa) Based on Mitochondrial CO1 Sequence Data

PubMed Central

Kitahara, Marcelo V.; Cairns, Stephen D.; Stolarski, Jarosław; Blair, David; Miller, David J.

2010-01-01

Background Classical morphological taxonomy places the approximately 1400 recognized species of Scleractinia (hard corals) into 27 families, but many aspects of coral evolution remain unclear despite the application of molecular phylogenetic methods. In part, this may be a consequence of such studies focusing on the reef-building (shallow water and zooxanthellate) Scleractinia, and largely ignoring the large number of deep-sea species. To better understand broad patterns of coral evolution, we generated molecular data for a broad and representative range of deep sea scleractinians collected off New Caledonia and Australia during the last decade, and conducted the most comprehensive molecular phylogenetic analysis to date of the order Scleractinia. Methodology Partial (595 bp) sequences of the mitochondrial cytochrome oxidase subunit 1 (CO1) gene were determined for 65 deep-sea (azooxanthellate) scleractinians and 11 shallow-water species. These new data were aligned with 158 published sequences, generating a 234 taxon dataset representing 25 of the 27 currently recognized scleractinian families. Principal Findings/Conclusions There was a striking discrepancy between the taxonomic validity of coral families consisting predominantly of deep-sea or shallow-water species. Most families composed predominantly of deep-sea azooxanthellate species were monophyletic in both maximum likelihood and Bayesian analyses but, by contrast (and consistent with previous studies), most families composed predominantly of shallow-water zooxanthellate taxa were polyphyletic, although Acroporidae, Poritidae, Pocilloporidae, and Fungiidae were exceptions to this general pattern. One factor contributing to this inconsistency may be the greater environmental stability of deep-sea environments, effectively removing taxonomic “noise” contributed by phenotypic plasticity. Our phylogenetic analyses imply that the most basal extant scleractinians are azooxanthellate solitary corals from deep-water, their divergence predating that of the robust and complex corals. Deep-sea corals are likely to be critical to understanding anthozoan evolution and the origins of the Scleractinia. PMID:20628613

Independent studies using deep sequencing resolve the same set of core bacterial species dominating gut communities of honey bees.

PubMed

Sabree, Zakee L; Hansen, Allison K; Moran, Nancy A

2012-01-01

Starting in 2003, numerous studies using culture-independent methodologies to characterize the gut microbiota of honey bees have retrieved a consistent and distinctive set of eight bacterial species, based on near identity of the 16S rRNA gene sequences. A recent study [Mattila HR, Rios D, Walker-Sperling VE, Roeselers G, Newton ILG (2012) Characterization of the active microbiotas associated with honey bees reveals healthier and broader communities when colonies are genetically diverse. PLoS ONE 7(3): e32962], using pyrosequencing of the V1-V2 hypervariable region of the 16S rRNA gene, reported finding entirely novel bacterial species in honey bee guts, and used taxonomic assignments from these reads to predict metabolic activities based on known metabolisms of cultivable species. To better understand this discrepancy, we analyzed the Mattila et al. pyrotag dataset. In contrast to the conclusions of Mattila et al., we found that the large majority of pyrotag sequences belonged to clusters for which representative sequences were identical to sequences from previously identified core species of the bee microbiota. On average, they represent 95% of the bacteria in each worker bee in the Mattila et al. dataset, a slightly lower value than that found in other studies. Some colonies contain small proportions of other bacteria, mostly species of Enterobacteriaceae. Reanalysis of the Mattila et al. dataset also did not support a relationship between abundances of Bifidobacterium and of putative pathogens or a significant difference in gut communities between colonies from queens that were singly or multiply mated. Additionally, consistent with previous studies, the dataset supports the occurrence of considerable strain variation within core species, even within single colonies. The roles of these bacteria within bees, or the implications of the strain variation, are not yet clear.

Impact of sequencing depth in ChIP-seq experiments

PubMed Central

Jung, Youngsook L.; Luquette, Lovelace J.; Ho, Joshua W.K.; Ferrari, Francesco; Tolstorukov, Michael; Minoda, Aki; Issner, Robbyn; Epstein, Charles B.; Karpen, Gary H.; Kuroda, Mitzi I.; Park, Peter J.

2014-01-01

In a chromatin immunoprecipitation followed by high-throughput sequencing (ChIP-seq) experiment, an important consideration in experimental design is the minimum number of sequenced reads required to obtain statistically significant results. We present an extensive evaluation of the impact of sequencing depth on identification of enriched regions for key histone modifications (H3K4me3, H3K36me3, H3K27me3 and H3K9me2/me3) using deep-sequenced datasets in human and fly. We propose to define sufficient sequencing depth as the number of reads at which detected enrichment regions increase <1% for an additional million reads. Although the required depth depends on the nature of the mark and the state of the cell in each experiment, we observe that sufficient depth is often reached at <20 million reads for fly. For human, there are no clear saturation points for the examined datasets, but our analysis suggests 40–50 million reads as a practical minimum for most marks. We also devise a mathematical model to estimate the sufficient depth and total genomic coverage of a mark. Lastly, we find that the five algorithms tested do not agree well for broad enrichment profiles, especially at lower depths. Our findings suggest that sufficient sequencing depth and an appropriate peak-calling algorithm are essential for ensuring robustness of conclusions derived from ChIP-seq data. PMID:24598259

Do pre-trained deep learning models improve computer-aided classification of digital mammograms?

NASA Astrophysics Data System (ADS)

Aboutalib, Sarah S.; Mohamed, Aly A.; Zuley, Margarita L.; Berg, Wendie A.; Luo, Yahong; Wu, Shandong

2018-02-01

Digital mammography screening is an important exam for the early detection of breast cancer and reduction in mortality. False positives leading to high recall rates, however, results in unnecessary negative consequences to patients and health care systems. In order to better aid radiologists, computer-aided tools can be utilized to improve distinction between image classifications and thus potentially reduce false recalls. The emergence of deep learning has shown promising results in the area of biomedical imaging data analysis. This study aimed to investigate deep learning and transfer learning methods that can improve digital mammography classification performance. In particular, we evaluated the effect of pre-training deep learning models with other imaging datasets in order to boost classification performance on a digital mammography dataset. Two types of datasets were used for pre-training: (1) a digitized film mammography dataset, and (2) a very large non-medical imaging dataset. By using either of these datasets to pre-train the network initially, and then fine-tuning with the digital mammography dataset, we found an increase in overall classification performance in comparison to a model without pre-training, with the very large non-medical dataset performing the best in improving the classification accuracy.

When less is more: 'slicing' sequencing data improves read decoding accuracy and de novo assembly quality.

PubMed

Lonardi, Stefano; Mirebrahim, Hamid; Wanamaker, Steve; Alpert, Matthew; Ciardo, Gianfranco; Duma, Denisa; Close, Timothy J

2015-09-15

As the invention of DNA sequencing in the 70s, computational biologists have had to deal with the problem of de novo genome assembly with limited (or insufficient) depth of sequencing. In this work, we investigate the opposite problem, that is, the challenge of dealing with excessive depth of sequencing. We explore the effect of ultra-deep sequencing data in two domains: (i) the problem of decoding reads to bacterial artificial chromosome (BAC) clones (in the context of the combinatorial pooling design we have recently proposed), and (ii) the problem of de novo assembly of BAC clones. Using real ultra-deep sequencing data, we show that when the depth of sequencing increases over a certain threshold, sequencing errors make these two problems harder and harder (instead of easier, as one would expect with error-free data), and as a consequence the quality of the solution degrades with more and more data. For the first problem, we propose an effective solution based on 'divide and conquer': we 'slice' a large dataset into smaller samples of optimal size, decode each slice independently, and then merge the results. Experimental results on over 15 000 barley BACs and over 4000 cowpea BACs demonstrate a significant improvement in the quality of the decoding and the final assembly. For the second problem, we show for the first time that modern de novo assemblers cannot take advantage of ultra-deep sequencing data. Python scripts to process slices and resolve decoding conflicts are available from http://goo.gl/YXgdHT; software Hashfilter can be downloaded from http://goo.gl/MIyZHs stelo@cs.ucr.edu or timothy.close@ucr.edu Supplementary data are available at Bioinformatics online. © The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.

Exploring Genomic Diversity Using Metagenomics of Deep-Sea Subsurface Microbes from the Louisville Seamount and the South Pacific Gyre

NASA Astrophysics Data System (ADS)

Tully, B. J.; Sylvan, J. B.; Heidelberg, J. F.; Huber, J. A.

2014-12-01

There are many limitations involved with sampling microbial diversity from deep-sea subsurface environments, ranging from physical sample collection, low microbial biomass, culturing at in situ conditions, and inefficient nucleic acid extractions. As such, we are continually modifying our methods to obtain better results and expanding what we know about microbes in these environments. Here we present analysis of metagenomes sequences from samples collected from 120 m within the Louisville Seamount and from the top 5-10cm of the sediment in the center of the south Pacific gyre (SPG). Both systems are low biomass with ~102 and ~104 cells per cm3 for Louisville Seamount samples analyzed and the SPG sediment, respectively. The Louisville Seamount represents the first in situ subseafloor basalt and the SPG sediments represent the first in situ low biomass sediment microbial metagenomes. Both of these environments, subseafloor basalt and sediments underlying oligotrophic ocean gyres, represent large provinces of the seafloor environment that remain understudied. Despite the low biomass and DNA generated from these samples, we have generated 16 near complete genomes (5 from Louisville and 11 from the SPG) from the two metagenomic datasets. These genomes are estimated to be between 51-100% complete and span a range of phylogenetic groups, including the Proteobacteria, Actinobacteria, Firmicutes, Chloroflexi, and unclassified bacterial groups. With these genomes, we have assessed potential functional capabilities of these organisms and performed a comparative analysis between the environmental genomes and previously sequenced relatives to determine possible adaptations that may elucidate survival mechanisms for these low energy environments. These methods illustrate a baseline analysis that can be applied to future metagenomic deep-sea subsurface datasets and will help to further our understanding of microbiology within these environments.

Joint deep shape and appearance learning: application to optic pathway glioma segmentation

NASA Astrophysics Data System (ADS)

Mansoor, Awais; Li, Ien; Packer, Roger J.; Avery, Robert A.; Linguraru, Marius George

2017-03-01

Automated tissue characterization is one of the major applications of computer-aided diagnosis systems. Deep learning techniques have recently demonstrated impressive performance for the image patch-based tissue characterization. However, existing patch-based tissue classification techniques struggle to exploit the useful shape information. Local and global shape knowledge such as the regional boundary changes, diameter, and volumetrics can be useful in classifying the tissues especially in scenarios where the appearance signature does not provide significant classification information. In this work, we present a deep neural network-based method for the automated segmentation of the tumors referred to as optic pathway gliomas (OPG) located within the anterior visual pathway (AVP; optic nerve, chiasm or tracts) using joint shape and appearance learning. Voxel intensity values of commonly used MRI sequences are generally not indicative of OPG. To be considered an OPG, current clinical practice dictates that some portion of AVP must demonstrate shape enlargement. The method proposed in this work integrates multiple sequence magnetic resonance image (T1, T2, and FLAIR) along with local boundary changes to train a deep neural network. For training and evaluation purposes, we used a dataset of multiple sequence MRI obtained from 20 subjects (10 controls, 10 NF1+OPG). To our best knowledge, this is the first deep representation learning-based approach designed to merge shape and multi-channel appearance data for the glioma detection. In our experiments, mean misclassification errors of 2:39% and 0:48% were observed respectively for glioma and control patches extracted from the AVP. Moreover, an overall dice similarity coefficient of 0:87+/-0:13 (0:93+/-0:06 for healthy tissue, 0:78+/-0:18 for glioma tissue) demonstrates the potential of the proposed method in the accurate localization and early detection of OPG.

Thermalnet: a Deep Convolutional Network for Synthetic Thermal Image Generation

NASA Astrophysics Data System (ADS)

Kniaz, V. V.; Gorbatsevich, V. S.; Mizginov, V. A.

2017-05-01

Deep convolutional neural networks have dramatically changed the landscape of the modern computer vision. Nowadays methods based on deep neural networks show the best performance among image recognition and object detection algorithms. While polishing of network architectures received a lot of scholar attention, from the practical point of view the preparation of a large image dataset for a successful training of a neural network became one of major challenges. This challenge is particularly profound for image recognition in wavelengths lying outside the visible spectrum. For example no infrared or radar image datasets large enough for successful training of a deep neural network are available to date in public domain. Recent advances of deep neural networks prove that they are also capable to do arbitrary image transformations such as super-resolution image generation, grayscale image colorisation and imitation of style of a given artist. Thus a natural question arise: how could be deep neural networks used for augmentation of existing large image datasets? This paper is focused on the development of the Thermalnet deep convolutional neural network for augmentation of existing large visible image datasets with synthetic thermal images. The Thermalnet network architecture is inspired by colorisation deep neural networks.

A New Approach for Mining Order-Preserving Submatrices Based on All Common Subsequences.

PubMed

Xue, Yun; Liao, Zhengling; Li, Meihang; Luo, Jie; Kuang, Qiuhua; Hu, Xiaohui; Li, Tiechen

2015-01-01

Order-preserving submatrices (OPSMs) have been applied in many fields, such as DNA microarray data analysis, automatic recommendation systems, and target marketing systems, as an important unsupervised learning model. Unfortunately, most existing methods are heuristic algorithms which are unable to reveal OPSMs entirely in NP-complete problem. In particular, deep OPSMs, corresponding to long patterns with few supporting sequences, incur explosive computational costs and are completely pruned by most popular methods. In this paper, we propose an exact method to discover all OPSMs based on frequent sequential pattern mining. First, an existing algorithm was adjusted to disclose all common subsequence (ACS) between every two row sequences, and therefore all deep OPSMs will not be missed. Then, an improved data structure for prefix tree was used to store and traverse ACS, and Apriori principle was employed to efficiently mine the frequent sequential pattern. Finally, experiments were implemented on gene and synthetic datasets. Results demonstrated the effectiveness and efficiency of this method.

MJO Signals in Latent Heating: Results from TRMM Retrievals

NASA Technical Reports Server (NTRS)

Zhang, Chidong; Ling, Jian; Hagos, Samson; Tao, Wei-Kuo; Lang, Steve; Takayabu, Yukari N.; Shige, Shoichi; Katsumata, Masaki; Olson, William S.; L'Ecuyer, Tristan

2010-01-01

The Madden-Julian Oscillation (MJO) is the dominant intraseasonal signal in the global tropical atmosphere. Almost all numerical climate models have difficulty to simulate realistic MJO. Four TRMM datasets of latent heating were diagnosed for signals in the MJO. In all four datasets, vertical structures of latent heating are dominated by two components, one deep with its peak above the melting level and one shallow with its peak below. Profiles of the two components are nearly ubiquitous in longitude, allowing a separation of the vertical and zonal/temporal variations when the latitudinal dependence is not considered. All four datasets exhibit robust MJO spectral signals in the deep component as eastward propagating spectral peaks centered at period of 50 days and zonal wavenumber 1, well distinguished from lower- and higher-frequency power and much stronger than the corresponding westward power. The shallow component shows similar but slightly less robust MJO spectral peaks. MJO signals were further extracted from a combination of band-pass (30 - 90 day) filtered deep and shallow components. Largest amplitudes of both deep and shallow components of the MJO are confined to the Indian and western Pacific Oceans. There is a local minimum in the deep components over the Maritime Continent. The shallow components of the MJO differ substantially among the four TRMM datasets in their detailed zonal distributions in the eastern hemisphere. In composites of the heating evolution through the life cycle of the MJO, the shallow components lead the deep ones in some datasets and at certain longitudes. In many respects, the four TRMM datasets agree well in their deep components, but not in their shallow components and the phase relations between the deep and shallow components. These results indicate that caution must be exercised in applications of these latent heating data.

Evolving Deep Networks Using HPC

DOE Office of Scientific and Technical Information (OSTI.GOV)

Young, Steven R.; Rose, Derek C.; Johnston, Travis

While a large number of deep learning networks have been studied and published that produce outstanding results on natural image datasets, these datasets only make up a fraction of those to which deep learning can be applied. These datasets include text data, audio data, and arrays of sensors that have very different characteristics than natural images. As these “best” networks for natural images have been largely discovered through experimentation and cannot be proven optimal on some theoretical basis, there is no reason to believe that they are the optimal network for these drastically different datasets. Hyperparameter search is thus oftenmore » a very important process when applying deep learning to a new problem. In this work we present an evolutionary approach to searching the possible space of network hyperparameters and construction that can scale to 18, 000 nodes. This approach is applied to datasets of varying types and characteristics where we demonstrate the ability to rapidly find best hyperparameters in order to enable practitioners to quickly iterate between idea and result.« less

Single-shot T2 mapping using overlapping-echo detachment planar imaging and a deep convolutional neural network.

PubMed

Cai, Congbo; Wang, Chao; Zeng, Yiqing; Cai, Shuhui; Liang, Dong; Wu, Yawen; Chen, Zhong; Ding, Xinghao; Zhong, Jianhui

2018-04-24

An end-to-end deep convolutional neural network (CNN) based on deep residual network (ResNet) was proposed to efficiently reconstruct reliable T 2 mapping from single-shot overlapping-echo detachment (OLED) planar imaging. The training dataset was obtained from simulations that were carried out on SPROM (Simulation with PRoduct Operator Matrix) software developed by our group. The relationship between the original OLED image containing two echo signals and the corresponding T 2 mapping was learned by ResNet training. After the ResNet was trained, it was applied to reconstruct the T 2 mapping from simulation and in vivo human brain data. Although the ResNet was trained entirely on simulated data, the trained network was generalized well to real human brain data. The results from simulation and in vivo human brain experiments show that the proposed method significantly outperforms the echo-detachment-based method. Reliable T 2 mapping with higher accuracy is achieved within 30 ms after the network has been trained, while the echo-detachment-based OLED reconstruction method took approximately 2 min. The proposed method will facilitate real-time dynamic and quantitative MR imaging via OLED sequence, and deep convolutional neural network has the potential to reconstruct maps from complex MRI sequences efficiently. © 2018 International Society for Magnetic Resonance in Medicine.

De novo characterization of the Chinese fir (Cunninghamia lanceolata) transcriptome and analysis of candidate genes involved in cellulose and lignin biosynthesis

PubMed Central

2012-01-01

Background Chinese fir (Cunninghamia lanceolata) is an important timber species that accounts for 20–30% of the total commercial timber production in China. However, the available genomic information of Chinese fir is limited, and this severely encumbers functional genomic analysis and molecular breeding in Chinese fir. Recently, major advances in transcriptome sequencing have provided fast and cost-effective approaches to generate large expression datasets that have proven to be powerful tools to profile the transcriptomes of non-model organisms with undetermined genomes. Results In this study, the transcriptomes of nine tissues from Chinese fir were analyzed using the Illumina HiSeq™ 2000 sequencing platform. Approximately 40 million paired-end reads were obtained, generating 3.62 gigabase pairs of sequencing data. These reads were assembled into 83,248 unique sequences (i.e. Unigenes) with an average length of 449 bp, amounting to 37.40 Mb. A total of 73,779 Unigenes were supported by more than 5 reads, 42,663 (57.83%) had homologs in the NCBI non-redundant and Swiss-Prot protein databases, corresponding to 27,224 unique protein entries. Of these Unigenes, 16,750 were assigned to Gene Ontology classes, and 14,877 were clustered into orthologous groups. A total of 21,689 (29.40%) were mapped to 119 pathways by BLAST comparison against the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. The majority of the genes encoding the enzymes in the biosynthetic pathways of cellulose and lignin were identified in the Unigene dataset by targeted searches of their annotations. And a number of candidate Chinese fir genes in the two metabolic pathways were discovered firstly. Eighteen genes related to cellulose and lignin biosynthesis were cloned for experimental validating of transcriptome data. Overall 49 Unigenes, covering different regions of these selected genes, were found by alignment. Their expression patterns in different tissues were analyzed by qRT-PCR to explore their putative functions. Conclusions A substantial fraction of transcript sequences was obtained from the deep sequencing of Chinese fir. The assembled Unigene dataset was used to discover candidate genes of cellulose and lignin biosynthesis. This transcriptome dataset will provide a comprehensive sequence resource for molecular genetics research of C. lanceolata. PMID:23171398

Deep neural networks show an equivalent and often superior performance to dermatologists in onychomycosis diagnosis: Automatic construction of onychomycosis datasets by region-based convolutional deep neural network.

PubMed

Han, Seung Seog; Park, Gyeong Hun; Lim, Woohyung; Kim, Myoung Shin; Na, Jung Im; Park, Ilwoo; Chang, Sung Eun

2018-01-01

Although there have been reports of the successful diagnosis of skin disorders using deep learning, unrealistically large clinical image datasets are required for artificial intelligence (AI) training. We created datasets of standardized nail images using a region-based convolutional neural network (R-CNN) trained to distinguish the nail from the background. We used R-CNN to generate training datasets of 49,567 images, which we then used to fine-tune the ResNet-152 and VGG-19 models. The validation datasets comprised 100 and 194 images from Inje University (B1 and B2 datasets, respectively), 125 images from Hallym University (C dataset), and 939 images from Seoul National University (D dataset). The AI (ensemble model; ResNet-152 + VGG-19 + feedforward neural networks) results showed test sensitivity/specificity/ area under the curve values of (96.0 / 94.7 / 0.98), (82.7 / 96.7 / 0.95), (92.3 / 79.3 / 0.93), (87.7 / 69.3 / 0.82) for the B1, B2, C, and D datasets. With a combination of the B1 and C datasets, the AI Youden index was significantly (p = 0.01) higher than that of 42 dermatologists doing the same assessment manually. For B1+C and B2+ D dataset combinations, almost none of the dermatologists performed as well as the AI. By training with a dataset comprising 49,567 images, we achieved a diagnostic accuracy for onychomycosis using deep learning that was superior to that of most of the dermatologists who participated in this study.

Cross-species inference of long non-coding RNAs greatly expands the ruminant transcriptome.

PubMed

Bush, Stephen J; Muriuki, Charity; McCulloch, Mary E B; Farquhar, Iseabail L; Clark, Emily L; Hume, David A

2018-04-24

mRNA-like long non-coding RNAs (lncRNAs) are a significant component of mammalian transcriptomes, although most are expressed only at low levels, with high tissue-specificity and/or at specific developmental stages. Thus, in many cases lncRNA detection by RNA-sequencing (RNA-seq) is compromised by stochastic sampling. To account for this and create a catalogue of ruminant lncRNAs, we compared de novo assembled lncRNAs derived from large RNA-seq datasets in transcriptional atlas projects for sheep and goats with previous lncRNAs assembled in cattle and human. We then combined the novel lncRNAs with the sheep transcriptional atlas to identify co-regulated sets of protein-coding and non-coding loci. Few lncRNAs could be reproducibly assembled from a single dataset, even with deep sequencing of the same tissues from multiple animals. Furthermore, there was little sequence overlap between lncRNAs that were assembled from pooled RNA-seq data. We combined positional conservation (synteny) with cross-species mapping of candidate lncRNAs to identify a consensus set of ruminant lncRNAs and then used the RNA-seq data to demonstrate detectable and reproducible expression in each species. In sheep, 20 to 30% of lncRNAs were located close to protein-coding genes with which they are strongly co-expressed, which is consistent with the evolutionary origin of some ncRNAs in enhancer sequences. Nevertheless, most of the lncRNAs are not co-expressed with neighbouring protein-coding genes. Alongside substantially expanding the ruminant lncRNA repertoire, the outcomes of our analysis demonstrate that stochastic sampling can be partly overcome by combining RNA-seq datasets from related species. This has practical implications for the future discovery of lncRNAs in other species.

Ultra Deep Sequencing of Listeria monocytogenes sRNA Transcriptome Revealed New Antisense RNAs

PubMed Central

Behrens, Sebastian; Widder, Stefanie; Mannala, Gopala Krishna; Qing, Xiaoxing; Madhugiri, Ramakanth; Kefer, Nathalie; Mraheil, Mobarak Abu; Rattei, Thomas; Hain, Torsten

2014-01-01

Listeria monocytogenes, a gram-positive pathogen, and causative agent of listeriosis, has become a widely used model organism for intracellular infections. Recent studies have identified small non-coding RNAs (sRNAs) as important factors for regulating gene expression and pathogenicity of L. monocytogenes. Increased speed and reduced costs of high throughput sequencing (HTS) techniques have made RNA sequencing (RNA-Seq) the state-of-the-art method to study bacterial transcriptomes. We created a large transcriptome dataset of L. monocytogenes containing a total of 21 million reads, using the SOLiD sequencing technology. The dataset contained cDNA sequences generated from L. monocytogenes RNA collected under intracellular and extracellular condition and additionally was size fractioned into three different size ranges from <40 nt, 40–150 nt and >150 nt. We report here, the identification of nine new sRNAs candidates of L. monocytogenes and a reevaluation of known sRNAs of L. monocytogenes EGD-e. Automatic comparison to known sRNAs revealed a high recovery rate of 55%, which was increased to 90% by manual revision of the data. Moreover, thorough classification of known sRNAs shed further light on their possible biological functions. Interestingly among the newly identified sRNA candidates are antisense RNAs (asRNAs) associated to the housekeeping genes purA, fumC and pgi and potentially their regulation, emphasizing the significance of sRNAs for metabolic adaptation in L. monocytogenes. PMID:24498259

A Comprehensive, Automatically Updated Fungal ITS Sequence Dataset for Reference-Based Chimera Control in Environmental Sequencing Efforts.

PubMed

Nilsson, R Henrik; Tedersoo, Leho; Ryberg, Martin; Kristiansson, Erik; Hartmann, Martin; Unterseher, Martin; Porter, Teresita M; Bengtsson-Palme, Johan; Walker, Donald M; de Sousa, Filipe; Gamper, Hannes Andres; Larsson, Ellen; Larsson, Karl-Henrik; Kõljalg, Urmas; Edgar, Robert C; Abarenkov, Kessy

2015-01-01

The nuclear ribosomal internal transcribed spacer (ITS) region is the most commonly chosen genetic marker for the molecular identification of fungi in environmental sequencing and molecular ecology studies. Several analytical issues complicate such efforts, one of which is the formation of chimeric-artificially joined-DNA sequences during PCR amplification or sequence assembly. Several software tools are currently available for chimera detection, but rely to various degrees on the presence of a chimera-free reference dataset for optimal performance. However, no such dataset is available for use with the fungal ITS region. This study introduces a comprehensive, automatically updated reference dataset for fungal ITS sequences based on the UNITE database for the molecular identification of fungi. This dataset supports chimera detection throughout the fungal kingdom and for full-length ITS sequences as well as partial (ITS1 or ITS2 only) datasets. The performance of the dataset on a large set of artificial chimeras was above 99.5%, and we subsequently used the dataset to remove nearly 1,000 compromised fungal ITS sequences from public circulation. The dataset is available at http://unite.ut.ee/repository.php and is subject to web-based third-party curation.

A Comprehensive, Automatically Updated Fungal ITS Sequence Dataset for Reference-Based Chimera Control in Environmental Sequencing Efforts

PubMed Central

Nilsson, R. Henrik; Tedersoo, Leho; Ryberg, Martin; Kristiansson, Erik; Hartmann, Martin; Unterseher, Martin; Porter, Teresita M.; Bengtsson-Palme, Johan; Walker, Donald M.; de Sousa, Filipe; Gamper, Hannes Andres; Larsson, Ellen; Larsson, Karl-Henrik; Kõljalg, Urmas; Edgar, Robert C.; Abarenkov, Kessy

2015-01-01

The nuclear ribosomal internal transcribed spacer (ITS) region is the most commonly chosen genetic marker for the molecular identification of fungi in environmental sequencing and molecular ecology studies. Several analytical issues complicate such efforts, one of which is the formation of chimeric—artificially joined—DNA sequences during PCR amplification or sequence assembly. Several software tools are currently available for chimera detection, but rely to various degrees on the presence of a chimera-free reference dataset for optimal performance. However, no such dataset is available for use with the fungal ITS region. This study introduces a comprehensive, automatically updated reference dataset for fungal ITS sequences based on the UNITE database for the molecular identification of fungi. This dataset supports chimera detection throughout the fungal kingdom and for full-length ITS sequences as well as partial (ITS1 or ITS2 only) datasets. The performance of the dataset on a large set of artificial chimeras was above 99.5%, and we subsequently used the dataset to remove nearly 1,000 compromised fungal ITS sequences from public circulation. The dataset is available at http://unite.ut.ee/repository.php and is subject to web-based third-party curation. PMID:25786896

DeepQA: improving the estimation of single protein model quality with deep belief networks.

PubMed

Cao, Renzhi; Bhattacharya, Debswapna; Hou, Jie; Cheng, Jianlin

2016-12-05

Protein quality assessment (QA) useful for ranking and selecting protein models has long been viewed as one of the major challenges for protein tertiary structure prediction. Especially, estimating the quality of a single protein model, which is important for selecting a few good models out of a large model pool consisting of mostly low-quality models, is still a largely unsolved problem. We introduce a novel single-model quality assessment method DeepQA based on deep belief network that utilizes a number of selected features describing the quality of a model from different perspectives, such as energy, physio-chemical characteristics, and structural information. The deep belief network is trained on several large datasets consisting of models from the Critical Assessment of Protein Structure Prediction (CASP) experiments, several publicly available datasets, and models generated by our in-house ab initio method. Our experiments demonstrate that deep belief network has better performance compared to Support Vector Machines and Neural Networks on the protein model quality assessment problem, and our method DeepQA achieves the state-of-the-art performance on CASP11 dataset. It also outperformed two well-established methods in selecting good outlier models from a large set of models of mostly low quality generated by ab initio modeling methods. DeepQA is a useful deep learning tool for protein single model quality assessment and protein structure prediction. The source code, executable, document and training/test datasets of DeepQA for Linux is freely available to non-commercial users at http://cactus.rnet.missouri.edu/DeepQA/ .

A Pilot Study of Biomedical Text Comprehension using an Attention-Based Deep Neural Reader: Design and Experimental Analysis

PubMed Central

Lee, Kyubum; Kim, Byounggun; Jeon, Minji; Kim, Jihye; Tan, Aik Choon

2018-01-01

Background With the development of artificial intelligence (AI) technology centered on deep-learning, the computer has evolved to a point where it can read a given text and answer a question based on the context of the text. Such a specific task is known as the task of machine comprehension. Existing machine comprehension tasks mostly use datasets of general texts, such as news articles or elementary school-level storybooks. However, no attempt has been made to determine whether an up-to-date deep learning-based machine comprehension model can also process scientific literature containing expert-level knowledge, especially in the biomedical domain. Objective This study aims to investigate whether a machine comprehension model can process biomedical articles as well as general texts. Since there is no dataset for the biomedical literature comprehension task, our work includes generating a large-scale question answering dataset using PubMed and manually evaluating the generated dataset. Methods We present an attention-based deep neural model tailored to the biomedical domain. To further enhance the performance of our model, we used a pretrained word vector and biomedical entity type embedding. We also developed an ensemble method of combining the results of several independent models to reduce the variance of the answers from the models. Results The experimental results showed that our proposed deep neural network model outperformed the baseline model by more than 7% on the new dataset. We also evaluated human performance on the new dataset. The human evaluation result showed that our deep neural model outperformed humans in comprehension by 22% on average. Conclusions In this work, we introduced a new task of machine comprehension in the biomedical domain using a deep neural model. Since there was no large-scale dataset for training deep neural models in the biomedical domain, we created the new cloze-style datasets Biomedical Knowledge Comprehension Title (BMKC_T) and Biomedical Knowledge Comprehension Last Sentence (BMKC_LS) (together referred to as BioMedical Knowledge Comprehension) using the PubMed corpus. The experimental results showed that the performance of our model is much higher than that of humans. We observed that our model performed consistently better regardless of the degree of difficulty of a text, whereas humans have difficulty when performing biomedical literature comprehension tasks that require expert level knowledge. PMID:29305341

A Two-Stream Deep Fusion Framework for High-Resolution Aerial Scene Classification

PubMed Central

Liu, Fuxian

2018-01-01

One of the challenging problems in understanding high-resolution remote sensing images is aerial scene classification. A well-designed feature representation method and classifier can improve classification accuracy. In this paper, we construct a new two-stream deep architecture for aerial scene classification. First, we use two pretrained convolutional neural networks (CNNs) as feature extractor to learn deep features from the original aerial image and the processed aerial image through saliency detection, respectively. Second, two feature fusion strategies are adopted to fuse the two different types of deep convolutional features extracted by the original RGB stream and the saliency stream. Finally, we use the extreme learning machine (ELM) classifier for final classification with the fused features. The effectiveness of the proposed architecture is tested on four challenging datasets: UC-Merced dataset with 21 scene categories, WHU-RS dataset with 19 scene categories, AID dataset with 30 scene categories, and NWPU-RESISC45 dataset with 45 challenging scene categories. The experimental results demonstrate that our architecture gets a significant classification accuracy improvement over all state-of-the-art references. PMID:29581722

A Two-Stream Deep Fusion Framework for High-Resolution Aerial Scene Classification.

PubMed

Yu, Yunlong; Liu, Fuxian

2018-01-01

One of the challenging problems in understanding high-resolution remote sensing images is aerial scene classification. A well-designed feature representation method and classifier can improve classification accuracy. In this paper, we construct a new two-stream deep architecture for aerial scene classification. First, we use two pretrained convolutional neural networks (CNNs) as feature extractor to learn deep features from the original aerial image and the processed aerial image through saliency detection, respectively. Second, two feature fusion strategies are adopted to fuse the two different types of deep convolutional features extracted by the original RGB stream and the saliency stream. Finally, we use the extreme learning machine (ELM) classifier for final classification with the fused features. The effectiveness of the proposed architecture is tested on four challenging datasets: UC-Merced dataset with 21 scene categories, WHU-RS dataset with 19 scene categories, AID dataset with 30 scene categories, and NWPU-RESISC45 dataset with 45 challenging scene categories. The experimental results demonstrate that our architecture gets a significant classification accuracy improvement over all state-of-the-art references.

ClimateNet: A Machine Learning dataset for Climate Science Research

NASA Astrophysics Data System (ADS)

Prabhat, M.; Biard, J.; Ganguly, S.; Ames, S.; Kashinath, K.; Kim, S. K.; Kahou, S.; Maharaj, T.; Beckham, C.; O'Brien, T. A.; Wehner, M. F.; Williams, D. N.; Kunkel, K.; Collins, W. D.

2017-12-01

Deep Learning techniques have revolutionized commercial applications in Computer vision, speech recognition and control systems. The key for all of these developments was the creation of a curated, labeled dataset ImageNet, for enabling multiple research groups around the world to develop methods, benchmark performance and compete with each other. The success of Deep Learning can be largely attributed to the broad availability of this dataset. Our empirical investigations have revealed that Deep Learning is similarly poised to benefit the task of pattern detection in climate science. Unfortunately, labeled datasets, a key pre-requisite for training, are hard to find. Individual research groups are typically interested in specialized weather patterns, making it hard to unify, and share datasets across groups and institutions. In this work, we are proposing ClimateNet: a labeled dataset that provides labeled instances of extreme weather patterns, as well as associated raw fields in model and observational output. We develop a schema in NetCDF to enumerate weather pattern classes/types, store bounding boxes, and pixel-masks. We are also working on a TensorFlow implementation to natively import such NetCDF datasets, and are providing a reference convolutional architecture for binary classification tasks. Our hope is that researchers in Climate Science, as well as ML/DL, will be able to use (and extend) ClimateNet to make rapid progress in the application of Deep Learning for Climate Science research.

Metatranscriptomic analysis of diverse microbial communities reveals core metabolic pathways and microbiome-specific functionality.

PubMed

Jiang, Yue; Xiong, Xuejian; Danska, Jayne; Parkinson, John

2016-01-12

Metatranscriptomics is emerging as a powerful technology for the functional characterization of complex microbial communities (microbiomes). Use of unbiased RNA-sequencing can reveal both the taxonomic composition and active biochemical functions of a complex microbial community. However, the lack of established reference genomes, computational tools and pipelines make analysis and interpretation of these datasets challenging. Systematic studies that compare data across microbiomes are needed to demonstrate the ability of such pipelines to deliver biologically meaningful insights on microbiome function. Here, we apply a standardized analytical pipeline to perform a comparative analysis of metatranscriptomic data from diverse microbial communities derived from mouse large intestine, cow rumen, kimchi culture, deep-sea thermal vent and permafrost. Sequence similarity searches allowed annotation of 19 to 76% of putative messenger RNA (mRNA) reads, with the highest frequency in the kimchi dataset due to its relatively low complexity and availability of closely related reference genomes. Metatranscriptomic datasets exhibited distinct taxonomic and functional signatures. From a metabolic perspective, we identified a common core of enzymes involved in amino acid, energy and nucleotide metabolism and also identified microbiome-specific pathways such as phosphonate metabolism (deep sea) and glycan degradation pathways (cow rumen). Integrating taxonomic and functional annotations within a novel visualization framework revealed the contribution of different taxa to metabolic pathways, allowing the identification of taxa that contribute unique functions. The application of a single, standard pipeline confirms that the rich taxonomic and functional diversity observed across microbiomes is not simply an artefact of different analysis pipelines but instead reflects distinct environmental influences. At the same time, our findings show how microbiome complexity and availability of reference genomes can impact comprehensive annotation of metatranscriptomes. Consequently, beyond the application of standardized pipelines, additional caution must be taken when interpreting their output and performing downstream, microbiome-specific, analyses. The pipeline used in these analyses along with a tutorial has been made freely available for download from our project website: http://www.compsysbio.org/microbiome .

Deep RNA-Seq to unlock the gene bank of floral development in Sinapis arvensis.

PubMed

Liu, Jia; Mei, Desheng; Li, Yunchang; Huang, Shunmou; Hu, Qiong

2014-01-01

Sinapis arvensis is a weed with strong biological activity. Despite being a problematic annual weed that contaminates agricultural crop yield, it is a valuable alien germplasm resource. It can be utilized for broadening the genetic background of Brassica crops with desirable agricultural traits like resistance to blackleg (Leptosphaeria maculans), stem rot (Sclerotinia sclerotium) and pod shatter (caused by FRUITFULL gene). However, few genetic studies of S. arvensis were reported because of the lack of genomic resources. In the present study, we performed de novo transcriptome sequencing to produce a comprehensive dataset for S. arvensis for the first time. We used Illumina paired-end sequencing technology to sequence the S. arvensis flower transcriptome and generated 40,981,443 reads that were assembled into 131,278 transcripts. We de novo assembled 96,562 high quality unigenes with an average length of 832 bp. A total of 33,662 full-length ORF complete sequences were identified, and 41,415 unigenes were mapped onto 128 pathways using the KEGG Pathway database. The annotated unigenes were compared against Brassica rapa, B. oleracea, B. napus and Arabidopsis thaliana. Among these unigenes, 76,324 were identified as putative homologs of annotated sequences in the public protein databases, of which 1194 were associated with plant hormone signal transduction and 113 were related to gibberellin homeostasis/signaling. Unigenes that did not match any of those sequence datasets were considered to be unique to S. arvensis. Furthermore, 21,321 simple sequence repeats were found. Our study will enhance the currently available resources for Brassicaceae and will provide a platform for future genomic studies for genetic improvement of Brassica crops.

Deep RNA-Seq to Unlock the Gene Bank of Floral Development in Sinapis arvensis

PubMed Central

Liu, Jia; Mei, Desheng; Li, Yunchang; Huang, Shunmou; Hu, Qiong

2014-01-01

Sinapis arvensis is a weed with strong biological activity. Despite being a problematic annual weed that contaminates agricultural crop yield, it is a valuable alien germplasm resource. It can be utilized for broadening the genetic background of Brassica crops with desirable agricultural traits like resistance to blackleg (Leptosphaeria maculans), stem rot (Sclerotinia sclerotium) and pod shatter (caused by FRUITFULL gene). However, few genetic studies of S. arvensis were reported because of the lack of genomic resources. In the present study, we performed de novo transcriptome sequencing to produce a comprehensive dataset for S. arvensis for the first time. We used Illumina paired-end sequencing technology to sequence the S. arvensis flower transcriptome and generated 40,981,443 reads that were assembled into 131,278 transcripts. We de novo assembled 96,562 high quality unigenes with an average length of 832 bp. A total of 33,662 full-length ORF complete sequences were identified, and 41,415 unigenes were mapped onto 128 pathways using the KEGG Pathway database. The annotated unigenes were compared against Brassica rapa, B. oleracea, B. napus and Arabidopsis thaliana. Among these unigenes, 76,324 were identified as putative homologs of annotated sequences in the public protein databases, of which 1194 were associated with plant hormone signal transduction and 113 were related to gibberellin homeostasis/signaling. Unigenes that did not match any of those sequence datasets were considered to be unique to S. arvensis. Furthermore, 21,321 simple sequence repeats were found. Our study will enhance the currently available resources for Brassicaceae and will provide a platform for future genomic studies for genetic improvement of Brassica crops. PMID:25192023

Comparing sequencing assays and human-machine analyses in actionable genomics for glioblastoma.

PubMed

Wrzeszczynski, Kazimierz O; Frank, Mayu O; Koyama, Takahiko; Rhrissorrakrai, Kahn; Robine, Nicolas; Utro, Filippo; Emde, Anne-Katrin; Chen, Bo-Juen; Arora, Kanika; Shah, Minita; Vacic, Vladimir; Norel, Raquel; Bilal, Erhan; Bergmann, Ewa A; Moore Vogel, Julia L; Bruce, Jeffrey N; Lassman, Andrew B; Canoll, Peter; Grommes, Christian; Harvey, Steve; Parida, Laxmi; Michelini, Vanessa V; Zody, Michael C; Jobanputra, Vaidehi; Royyuru, Ajay K; Darnell, Robert B

2017-08-01

To analyze a glioblastoma tumor specimen with 3 different platforms and compare potentially actionable calls from each. Tumor DNA was analyzed by a commercial targeted panel. In addition, tumor-normal DNA was analyzed by whole-genome sequencing (WGS) and tumor RNA was analyzed by RNA sequencing (RNA-seq). The WGS and RNA-seq data were analyzed by a team of bioinformaticians and cancer oncologists, and separately by IBM Watson Genomic Analytics (WGA), an automated system for prioritizing somatic variants and identifying drugs. More variants were identified by WGS/RNA analysis than by targeted panels. WGA completed a comparable analysis in a fraction of the time required by the human analysts. The development of an effective human-machine interface in the analysis of deep cancer genomic datasets may provide potentially clinically actionable calls for individual patients in a more timely and efficient manner than currently possible. NCT02725684.

Identification of Free-Living and Particle-Associated Microbial Communities Present in Hadal Regions of the Mariana Trench.

PubMed

Tarn, Jonathan; Peoples, Logan M; Hardy, Kevin; Cameron, James; Bartlett, Douglas H

2016-01-01

Relatively few studies have described the microbial populations present in ultra-deep hadal environments, largely as a result of difficulties associated with sampling. Here we report Illumina-tag V6 16S rRNA sequence-based analyses of the free-living and particle-associated microbial communities recovered from locations within two of the deepest hadal sites on Earth, the Challenger Deep (10,918 meters below surface-mbs) and the Sirena Deep (10,667 mbs) within the Mariana Trench, as well as one control site (Ulithi Atoll, 761 mbs). Seawater samples were collected using an autonomous lander positioned ~1 m above the seafloor. The bacterial populations within the Mariana Trench bottom water samples were dissimilar to other deep-sea microbial communities, though with overlap with those of diffuse flow hydrothermal vents and deep-subsurface locations. Distinct particle-associated and free-living bacterial communities were found to exist. The hadal bacterial populations were also markedly different from one another, indicating the likelihood of different chemical conditions at the two sites. In contrast to the bacteria, the hadal archaeal communities were more similar to other less deep datasets and to each other due to an abundance of cosmopolitan deep-sea taxa. The hadal communities were enriched in 34 bacterial and 4 archaeal operational taxonomic units (OTUs) including members of the Gammaproteobacteria, Epsilonproteobacteria, Marinimicrobia, Cyanobacteria, Deltaproteobacteria, Gemmatimonadetes, Atribacteria, Spirochaetes, and Euryarchaeota. Sequences matching cultivated piezophiles were notably enriched in the Challenger Deep, especially within the particle-associated fraction, and were found in higher abundances than in other hadal studies, where they were either far less prevalent or missing. Our results indicate the importance of heterotrophy, sulfur-cycling, and methane and hydrogen utilization within the bottom waters of the deeper regions of the Mariana Trench, and highlight novel community features of these extreme habitats.

Identification of Free-Living and Particle-Associated Microbial Communities Present in Hadal Regions of the Mariana Trench

PubMed Central

Tarn, Jonathan; Peoples, Logan M.; Hardy, Kevin; Cameron, James; Bartlett, Douglas H.

2016-01-01

Relatively few studies have described the microbial populations present in ultra-deep hadal environments, largely as a result of difficulties associated with sampling. Here we report Illumina-tag V6 16S rRNA sequence-based analyses of the free-living and particle-associated microbial communities recovered from locations within two of the deepest hadal sites on Earth, the Challenger Deep (10,918 meters below surface-mbs) and the Sirena Deep (10,667 mbs) within the Mariana Trench, as well as one control site (Ulithi Atoll, 761 mbs). Seawater samples were collected using an autonomous lander positioned ~1 m above the seafloor. The bacterial populations within the Mariana Trench bottom water samples were dissimilar to other deep-sea microbial communities, though with overlap with those of diffuse flow hydrothermal vents and deep-subsurface locations. Distinct particle-associated and free-living bacterial communities were found to exist. The hadal bacterial populations were also markedly different from one another, indicating the likelihood of different chemical conditions at the two sites. In contrast to the bacteria, the hadal archaeal communities were more similar to other less deep datasets and to each other due to an abundance of cosmopolitan deep-sea taxa. The hadal communities were enriched in 34 bacterial and 4 archaeal operational taxonomic units (OTUs) including members of the Gammaproteobacteria, Epsilonproteobacteria, Marinimicrobia, Cyanobacteria, Deltaproteobacteria, Gemmatimonadetes, Atribacteria, Spirochaetes, and Euryarchaeota. Sequences matching cultivated piezophiles were notably enriched in the Challenger Deep, especially within the particle-associated fraction, and were found in higher abundances than in other hadal studies, where they were either far less prevalent or missing. Our results indicate the importance of heterotrophy, sulfur-cycling, and methane and hydrogen utilization within the bottom waters of the deeper regions of the Mariana Trench, and highlight novel community features of these extreme habitats. PMID:27242695

Human fatigue expression recognition through image-based dynamic multi-information and bimodal deep learning

NASA Astrophysics Data System (ADS)

Zhao, Lei; Wang, Zengcai; Wang, Xiaojin; Qi, Yazhou; Liu, Qing; Zhang, Guoxin

2016-09-01

Human fatigue is an important cause of traffic accidents. To improve the safety of transportation, we propose, in this paper, a framework for fatigue expression recognition using image-based facial dynamic multi-information and a bimodal deep neural network. First, the landmark of face region and the texture of eye region, which complement each other in fatigue expression recognition, are extracted from facial image sequences captured by a single camera. Then, two stacked autoencoder neural networks are trained for landmark and texture, respectively. Finally, the two trained neural networks are combined by learning a joint layer on top of them to construct a bimodal deep neural network. The model can be used to extract a unified representation that fuses landmark and texture modalities together and classify fatigue expressions accurately. The proposed system is tested on a human fatigue dataset obtained from an actual driving environment. The experimental results demonstrate that the proposed method performs stably and robustly, and that the average accuracy achieves 96.2%.

The Hubble Deep UV Legacy Survey (HDUV)

NASA Astrophysics Data System (ADS)

Montes, Mireia; Oesch, Pascal

2015-08-01

Deep HST imaging has shown that the overall star formation density and UV light density at z>3 is dominated by faint, blue galaxies. Remarkably, very little is known about the equivalent galaxy population at lower redshifts. Understanding how these galaxies evolve across the epoch of peak cosmic star-formation is key to a complete picture of galaxy evolution. Here, we present a new HST WFC3/UVIS program, the Hubble Deep UV (HDUV) legacy survey. The HDUV is a 132 orbit program to obtain deep imaging in two filters (F275W and F336W) over the two CANDELS Deep fields. We will cover ~100 arcmin2 sampling the rest-frame far-UV at z>~0.5, this will provide a unique legacy dataset with exquisite HST multi-wavelength imaging as well as ancillary HST grism NIR spectroscopy for a detailed study of faint, star-forming galaxies at z~0.5-2. The HDUV will enable a wealth of research by the community, which includes tracing the evolution of the FUV luminosity function over the peak of the star formation rate density from z~3 down to z~0.5, measuring the physical properties of sub-L* galaxies, and characterizing resolved stellar populations to decipher the build-up of the Hubble sequence from sub-galactic clumps. This poster provides an overview of the HDUV survey and presents the reduced data products and catalogs which will be released to the community, reaching down to 27.5-28.0 mag at 5 sigma. By directly sampling the rest-frame far-UV at z>~0.5, this will provide a unique legacy dataset with exquisite HST multi-wavelength imaging as well as ancillary HST grism NIR spectroscopy for a detailed study of faint, star-forming galaxies at z~0.5-2. The HDUV will enable a wealth of research by the community, which includes tracing the evolution of the FUV luminosity function over the peak of the star formation rate density from z~3 down to z~0.5, measuring the physical properties of sub-L* galaxies, and characterizing resolved stellar populations to decipher the build-up of the Hubble sequence from sub-galactic clumps. This poster provides an overview of the HDUV survey and presents reduced data products and catalogs which will be released to the community.

Recurrent neural networks for breast lesion classification based on DCE-MRIs

NASA Astrophysics Data System (ADS)

Antropova, Natasha; Huynh, Benjamin; Giger, Maryellen

2018-02-01

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) plays a significant role in breast cancer screening, cancer staging, and monitoring response to therapy. Recently, deep learning methods are being rapidly incorporated in image-based breast cancer diagnosis and prognosis. However, most of the current deep learning methods make clinical decisions based on 2-dimentional (2D) or 3D images and are not well suited for temporal image data. In this study, we develop a deep learning methodology that enables integration of clinically valuable temporal components of DCE-MRIs into deep learning-based lesion classification. Our work is performed on a database of 703 DCE-MRI cases for the task of distinguishing benign and malignant lesions, and uses the area under the ROC curve (AUC) as the performance metric in conducting that task. We train a recurrent neural network, specifically a long short-term memory network (LSTM), on sequences of image features extracted from the dynamic MRI sequences. These features are extracted with VGGNet, a convolutional neural network pre-trained on a large dataset of natural images ImageNet. The features are obtained from various levels of the network, to capture low-, mid-, and high-level information about the lesion. Compared to a classification method that takes as input only images at a single time-point (yielding an AUC = 0.81 (se = 0.04)), our LSTM method improves lesion classification with an AUC of 0.85 (se = 0.03).

Wave equation datuming applied to marine OBS data and to land high resolution seismic profiling

NASA Astrophysics Data System (ADS)

Barison, Erika; Brancatelli, Giuseppe; Nicolich, Rinaldo; Accaino, Flavio; Giustiniani, Michela; Tinivella, Umberta

2011-03-01

One key step in seismic data processing flows is the computation of static corrections, which relocate shots and receivers at the same datum plane and remove near surface weathering effects. We applied a standard static correction and a wave equation datuming and compared the obtained results in two case studies: 1) a sparse ocean bottom seismometers dataset for deep crustal prospecting; 2) a high resolution land reflection dataset for hydrogeological investigation. In both cases, a detailed velocity field, obtained by tomographic inversion of the first breaks, was adopted to relocate shots and receivers to the datum plane. The results emphasize the importance of wave equation datuming to properly handle complex near surface conditions. In the first dataset, the deployed ocean bottom seismometers were relocated to the sea level (shot positions) and a standard processing sequence was subsequently applied to the output. In the second dataset, the application of wave equation datuming allowed us to remove the coherent noise, such as ground roll, and to improve the image quality with respect to the application of static correction. The comparison of the two approaches evidences that the main reflecting markers are better resolved when the wave equation datuming procedure is adopted.

A hybrid deep learning approach to predict malignancy of breast lesions using mammograms

NASA Astrophysics Data System (ADS)

Wang, Yunzhi; Heidari, Morteza; Mirniaharikandehei, Seyedehnafiseh; Gong, Jing; Qian, Wei; Qiu, Yuchen; Zheng, Bin

2018-03-01

Applying deep learning technology to medical imaging informatics field has been recently attracting extensive research interest. However, the limited medical image dataset size often reduces performance and robustness of the deep learning based computer-aided detection and/or diagnosis (CAD) schemes. In attempt to address this technical challenge, this study aims to develop and evaluate a new hybrid deep learning based CAD approach to predict likelihood of a breast lesion detected on mammogram being malignant. In this approach, a deep Convolutional Neural Network (CNN) was firstly pre-trained using the ImageNet dataset and serve as a feature extractor. A pseudo-color Region of Interest (ROI) method was used to generate ROIs with RGB channels from the mammographic images as the input to the pre-trained deep network. The transferred CNN features from different layers of the CNN were then obtained and a linear support vector machine (SVM) was trained for the prediction task. By applying to a dataset involving 301 suspicious breast lesions and using a leave-one-case-out validation method, the areas under the ROC curves (AUC) = 0.762 and 0.792 using the traditional CAD scheme and the proposed deep learning based CAD scheme, respectively. An ensemble classifier that combines the classification scores generated by the two schemes yielded an improved AUC value of 0.813. The study results demonstrated feasibility and potentially improved performance of applying a new hybrid deep learning approach to develop CAD scheme using a relatively small dataset of medical images.

Interpretable dimensionality reduction of single cell transcriptome data with deep generative models.

PubMed

Ding, Jiarui; Condon, Anne; Shah, Sohrab P

2018-05-21

Single-cell RNA-sequencing has great potential to discover cell types, identify cell states, trace development lineages, and reconstruct the spatial organization of cells. However, dimension reduction to interpret structure in single-cell sequencing data remains a challenge. Existing algorithms are either not able to uncover the clustering structures in the data or lose global information such as groups of clusters that are close to each other. We present a robust statistical model, scvis, to capture and visualize the low-dimensional structures in single-cell gene expression data. Simulation results demonstrate that low-dimensional representations learned by scvis preserve both the local and global neighbor structures in the data. In addition, scvis is robust to the number of data points and learns a probabilistic parametric mapping function to add new data points to an existing embedding. We then use scvis to analyze four single-cell RNA-sequencing datasets, exemplifying interpretable two-dimensional representations of the high-dimensional single-cell RNA-sequencing data.

Bayesian mixture analysis for metagenomic community profiling.

PubMed

Morfopoulou, Sofia; Plagnol, Vincent

2015-09-15

Deep sequencing of clinical samples is now an established tool for the detection of infectious pathogens, with direct medical applications. The large amount of data generated produces an opportunity to detect species even at very low levels, provided that computational tools can effectively profile the relevant metagenomic communities. Data interpretation is complicated by the fact that short sequencing reads can match multiple organisms and by the lack of completeness of existing databases, in particular for viral pathogens. Here we present metaMix, a Bayesian mixture model framework for resolving complex metagenomic mixtures. We show that the use of parallel Monte Carlo Markov chains for the exploration of the species space enables the identification of the set of species most likely to contribute to the mixture. We demonstrate the greater accuracy of metaMix compared with relevant methods, particularly for profiling complex communities consisting of several related species. We designed metaMix specifically for the analysis of deep transcriptome sequencing datasets, with a focus on viral pathogen detection; however, the principles are generally applicable to all types of metagenomic mixtures. metaMix is implemented as a user friendly R package, freely available on CRAN: http://cran.r-project.org/web/packages/metaMix sofia.morfopoulou.10@ucl.ac.uk Supplementary data are available at Bionformatics online. © The Author 2015. Published by Oxford University Press.

A Pilot Study of Biomedical Text Comprehension using an Attention-Based Deep Neural Reader: Design and Experimental Analysis.

PubMed

Kim, Seongsoon; Park, Donghyeon; Choi, Yonghwa; Lee, Kyubum; Kim, Byounggun; Jeon, Minji; Kim, Jihye; Tan, Aik Choon; Kang, Jaewoo

2018-01-05

With the development of artificial intelligence (AI) technology centered on deep-learning, the computer has evolved to a point where it can read a given text and answer a question based on the context of the text. Such a specific task is known as the task of machine comprehension. Existing machine comprehension tasks mostly use datasets of general texts, such as news articles or elementary school-level storybooks. However, no attempt has been made to determine whether an up-to-date deep learning-based machine comprehension model can also process scientific literature containing expert-level knowledge, especially in the biomedical domain. This study aims to investigate whether a machine comprehension model can process biomedical articles as well as general texts. Since there is no dataset for the biomedical literature comprehension task, our work includes generating a large-scale question answering dataset using PubMed and manually evaluating the generated dataset. We present an attention-based deep neural model tailored to the biomedical domain. To further enhance the performance of our model, we used a pretrained word vector and biomedical entity type embedding. We also developed an ensemble method of combining the results of several independent models to reduce the variance of the answers from the models. The experimental results showed that our proposed deep neural network model outperformed the baseline model by more than 7% on the new dataset. We also evaluated human performance on the new dataset. The human evaluation result showed that our deep neural model outperformed humans in comprehension by 22% on average. In this work, we introduced a new task of machine comprehension in the biomedical domain using a deep neural model. Since there was no large-scale dataset for training deep neural models in the biomedical domain, we created the new cloze-style datasets Biomedical Knowledge Comprehension Title (BMKC_T) and Biomedical Knowledge Comprehension Last Sentence (BMKC_LS) (together referred to as BioMedical Knowledge Comprehension) using the PubMed corpus. The experimental results showed that the performance of our model is much higher than that of humans. We observed that our model performed consistently better regardless of the degree of difficulty of a text, whereas humans have difficulty when performing biomedical literature comprehension tasks that require expert level knowledge. ©Seongsoon Kim, Donghyeon Park, Yonghwa Choi, Kyubum Lee, Byounggun Kim, Minji Jeon, Jihye Kim, Aik Choon Tan, Jaewoo Kang. Originally published in JMIR Medical Informatics (http://medinform.jmir.org), 05.01.2018.

MEBS, a software platform to evaluate large (meta)genomic collections according to their metabolic machinery: unraveling the sulfur cycle

PubMed Central

Zapata-Peñasco, Icoquih; Poot-Hernandez, Augusto Cesar; Eguiarte, Luis E

2017-01-01

Abstract The increasing number of metagenomic and genomic sequences has dramatically improved our understanding of microbial diversity, yet our ability to infer metabolic capabilities in such datasets remains challenging. We describe the Multigenomic Entropy Based Score pipeline (MEBS), a software platform designed to evaluate, compare, and infer complex metabolic pathways in large “omic” datasets, including entire biogeochemical cycles. MEBS is open source and available through https://github.com/eead-csic-compbio/metagenome_Pfam_score. To demonstrate its use, we modeled the sulfur cycle by exhaustively curating the molecular and ecological elements involved (compounds, genes, metabolic pathways, and microbial taxa). This information was reduced to a collection of 112 characteristic Pfam protein domains and a list of complete-sequenced sulfur genomes. Using the mathematical framework of relative entropy (H΄), we quantitatively measured the enrichment of these domains among sulfur genomes. The entropy of each domain was used both to build up a final score that indicates whether a (meta)genomic sample contains the metabolic machinery of interest and to propose marker domains in metagenomic sequences such as DsrC (PF04358). MEBS was benchmarked with a dataset of 2107 non-redundant microbial genomes from RefSeq and 935 metagenomes from MG-RAST. Its performance, reproducibility, and robustness were evaluated using several approaches, including random sampling, linear regression models, receiver operator characteristic plots, and the area under the curve metric (AUC). Our results support the broad applicability of this algorithm to accurately classify (AUC = 0.985) hard-to-culture genomes (e.g., Candidatus Desulforudis audaxviator), previously characterized ones, and metagenomic environments such as hydrothermal vents, or deep-sea sediment. Our benchmark indicates that an entropy-based score can capture the metabolic machinery of interest and can be used to efficiently classify large genomic and metagenomic datasets, including uncultivated/unexplored taxa. PMID:29069412

MEBS, a software platform to evaluate large (meta)genomic collections according to their metabolic machinery: unraveling the sulfur cycle.

PubMed

De Anda, Valerie; Zapata-Peñasco, Icoquih; Poot-Hernandez, Augusto Cesar; Eguiarte, Luis E; Contreras-Moreira, Bruno; Souza, Valeria

2017-11-01

The increasing number of metagenomic and genomic sequences has dramatically improved our understanding of microbial diversity, yet our ability to infer metabolic capabilities in such datasets remains challenging. We describe the Multigenomic Entropy Based Score pipeline (MEBS), a software platform designed to evaluate, compare, and infer complex metabolic pathways in large "omic" datasets, including entire biogeochemical cycles. MEBS is open source and available through https://github.com/eead-csic-compbio/metagenome_Pfam_score. To demonstrate its use, we modeled the sulfur cycle by exhaustively curating the molecular and ecological elements involved (compounds, genes, metabolic pathways, and microbial taxa). This information was reduced to a collection of 112 characteristic Pfam protein domains and a list of complete-sequenced sulfur genomes. Using the mathematical framework of relative entropy (H΄), we quantitatively measured the enrichment of these domains among sulfur genomes. The entropy of each domain was used both to build up a final score that indicates whether a (meta)genomic sample contains the metabolic machinery of interest and to propose marker domains in metagenomic sequences such as DsrC (PF04358). MEBS was benchmarked with a dataset of 2107 non-redundant microbial genomes from RefSeq and 935 metagenomes from MG-RAST. Its performance, reproducibility, and robustness were evaluated using several approaches, including random sampling, linear regression models, receiver operator characteristic plots, and the area under the curve metric (AUC). Our results support the broad applicability of this algorithm to accurately classify (AUC = 0.985) hard-to-culture genomes (e.g., Candidatus Desulforudis audaxviator), previously characterized ones, and metagenomic environments such as hydrothermal vents, or deep-sea sediment. Our benchmark indicates that an entropy-based score can capture the metabolic machinery of interest and can be used to efficiently classify large genomic and metagenomic datasets, including uncultivated/unexplored taxa. © The Author 2017. Published by Oxford University Press.

High resolution depth reconstruction from monocular images and sparse point clouds using deep convolutional neural network

NASA Astrophysics Data System (ADS)

Dimitrievski, Martin; Goossens, Bart; Veelaert, Peter; Philips, Wilfried

2017-09-01

Understanding the 3D structure of the environment is advantageous for many tasks in the field of robotics and autonomous vehicles. From the robot's point of view, 3D perception is often formulated as a depth image reconstruction problem. In the literature, dense depth images are often recovered deterministically from stereo image disparities. Other systems use an expensive LiDAR sensor to produce accurate, but semi-sparse depth images. With the advent of deep learning there have also been attempts to estimate depth by only using monocular images. In this paper we combine the best of the two worlds, focusing on a combination of monocular images and low cost LiDAR point clouds. We explore the idea that very sparse depth information accurately captures the global scene structure while variations in image patches can be used to reconstruct local depth to a high resolution. The main contribution of this paper is a supervised learning depth reconstruction system based on a deep convolutional neural network. The network is trained on RGB image patches reinforced with sparse depth information and the output is a depth estimate for each pixel. Using image and point cloud data from the KITTI vision dataset we are able to learn a correspondence between local RGB information and local depth, while at the same time preserving the global scene structure. Our results are evaluated on sequences from the KITTI dataset and our own recordings using a low cost camera and LiDAR setup.

Deep learning in breast cancer risk assessment: evaluation of convolutional neural networks on a clinical dataset of full-field digital mammograms.

PubMed

Li, Hui; Giger, Maryellen L; Huynh, Benjamin Q; Antropova, Natalia O

2017-10-01

To evaluate deep learning in the assessment of breast cancer risk in which convolutional neural networks (CNNs) with transfer learning are used to extract parenchymal characteristics directly from full-field digital mammographic (FFDM) images instead of using computerized radiographic texture analysis (RTA), 456 clinical FFDM cases were included: a "high-risk" BRCA1/2 gene-mutation carriers dataset (53 cases), a "high-risk" unilateral cancer patients dataset (75 cases), and a "low-risk dataset" (328 cases). Deep learning was compared to the use of features from RTA, as well as to a combination of both in the task of distinguishing between high- and low-risk subjects. Similar classification performances were obtained using CNN [area under the curve [Formula: see text]; standard error [Formula: see text

Light in the darkness: New perspective on lanternfish relationships and classification using genomic and morphological data.

PubMed

Martin, Rene P; Olson, Emily E; Girard, Matthew G; Smith, Wm Leo; Davis, Matthew P

2018-04-01

Massive parallel sequencing allows scientists to gather DNA sequences composed of millions of base pairs that can be combined into large datasets and analyzed to infer organismal relationships at a genome-wide scale in non-model organisms. Although the use of these large datasets is becoming more widespread, little to no work has been done in estimating phylogenetic relationships using UCEs in deep-sea fishes. Among deep-sea animals, the 257 species of lanternfishes (Myctophiformes) are among the most important open-ocean lineages, representing half of all mesopelagic vertebrate biomass. With this relative abundance, they are key members of the midwater food web where they feed on smaller invertebrates and fishes in addition to being a primary prey item for other open-ocean animals. Understanding the evolution and relationships of midwater organisms generally, and this dominant group of fishes in particular, is necessary for understanding and preserving the underexplored deep-sea ecosystem. Despite substantial congruence in the evolutionary relationships among deep-sea lanternfishes at higher classification levels in previous studies, the relationships among tribes, genera, and species within Myctophidae often conflict across phylogenetic studies or lack resolution and support. Herein we provide the first genome-scale phylogenetic analysis of lanternfishes, and we integrate these data from across the nuclear genome with additional protein-coding gene sequences and morphological data to further test evolutionary relationships among lanternfishes. Our phylogenetic hypotheses of relationships among lanternfishes are entirely congruent across a diversity of analyses that vary in methods, taxonomic sampling, and data analyzed. Within the Myctophiformes, the Neoscopelidae is inferred to be monophyletic and sister to a monophyletic Myctophidae. The current classification of lanternfishes is incongruent with our phylogenetic tree, so we recommend revisions that retain much of the traditional tribal structure and recognize five subfamilies instead of the traditional two subfamilies. The revised monophyletic taxonomy of myctophids includes the elevation of three former lampanyctine tribes to subfamilies. A restricted Lampanyctinae was recovered sister to Notolychninae. These two clades together were recovered as the sister group to the Gymnoscopelinae. Combined, these three subfamilies were recovered as the sister group to a clade composed of a monophyletic Diaphinae sister to the traditional Myctophinae. Our results corroborate recent multilocus molecular studies that infer a polyphyletic Myctophum in Myctophinae, and a para- or polyphyletic Lampanyctus and Nannobrachium within Lampanyctinae. We resurrect Dasyscopelus and Ctenoscopelus for the independent clades traditionally classified as species of Myctophum, and we place Nannobrachium into the synonymy of Lampanyctus. Copyright © 2017 Elsevier Inc. All rights reserved.

Deep sequencing of small RNA repertoires in mice reveals metabolic disorders-associated hepatic miRNAs.

PubMed

Liang, Tingming; Liu, Chang; Ye, Zhenchao

2013-01-01

Obesity and associated metabolic disorders contribute importantly to the metabolic syndrome. On the other hand, microRNAs (miRNAs) are a class of small non-coding RNAs that repress target gene expression by inducing mRNA degradation and/or translation repression. Dysregulation of specific miRNAs in obesity may influence energy metabolism and cause insulin resistance, which leads to dyslipidemia, steatosis hepatis and type 2 diabetes. In the present study, we comprehensively analyzed and validated dysregulated miRNAs in ob/ob mouse liver, as well as miRNA groups based on miRNA gene cluster and gene family by using deep sequencing miRNA datasets. We found that over 13.8% of the total analyzed miRNAs were dysregulated, of which 37 miRNA species showed significantly differential expression. Further RT-qPCR analysis in some selected miRNAs validated the similar expression patterns observed in deep sequencing. Interestingly, we found that miRNA gene cluster and family always showed consistent dysregulation patterns in ob/ob mouse liver, although they had various enrichment levels. Functional enrichment analysis revealed the versatile physiological roles (over six signal pathways and five human diseases) of these miRNAs. Biological studies indicated that overexpression of miR-126 or inhibition of miR-24 in AML-12 cells attenuated free fatty acids-induced fat accumulation. Taken together, our data strongly suggest that obesity and metabolic disturbance are tightly associated with functional miRNAs. We also identified hepatic miRNA candidates serving as potential biomarkers for the diagnose of the metabolic syndrome.

Genome-wide assessment of differential translations with ribosome profiling data.

PubMed

Xiao, Zhengtao; Zou, Qin; Liu, Yu; Yang, Xuerui

2016-04-04

The closely regulated process of mRNA translation is crucial for precise control of protein abundance and quality. Ribosome profiling, a combination of ribosome foot-printing and RNA deep sequencing, has been used in a large variety of studies to quantify genome-wide mRNA translation. Here, we developed Xtail, an analysis pipeline tailored for ribosome profiling data that comprehensively and accurately identifies differentially translated genes in pairwise comparisons. Applied on simulated and real datasets, Xtail exhibits high sensitivity with minimal false-positive rates, outperforming existing methods in the accuracy of quantifying differential translations. With published ribosome profiling datasets, Xtail does not only reveal differentially translated genes that make biological sense, but also uncovers new events of differential translation in human cancer cells on mTOR signalling perturbation and in human primary macrophages on interferon gamma (IFN-γ) treatment. This demonstrates the value of Xtail in providing novel insights into the molecular mechanisms that involve translational dysregulations.

Meraculous: De Novo Genome Assembly with Short Paired-End Reads

DOE Office of Scientific and Technical Information (OSTI.GOV)

Chapman, Jarrod A.; Ho, Isaac; Sunkara, Sirisha

2011-08-18

We describe a new algorithm, meraculous, for whole genome assembly of deep paired-end short reads, and apply it to the assembly of a dataset of paired 75-bp Illumina reads derived from the 15.4 megabase genome of the haploid yeast Pichia stipitis. More than 95% of the genome is recovered, with no errors; half the assembled sequence is in contigs longer than 101 kilobases and in scaffolds longer than 269 kilobases. Incorporating fosmid ends recovers entire chromosomes. Meraculous relies on an efficient and conservative traversal of the subgraph of the k-mer (deBruijn) graph of oligonucleotides with unique high quality extensions inmore » the dataset, avoiding an explicit error correction step as used in other short-read assemblers. A novel memory-efficient hashing scheme is introduced. The resulting contigs are ordered and oriented using paired reads separated by ~280 bp or ~3.2 kbp, and many gaps between contigs can be closed using paired-end placements. Practical issues with the dataset are described, and prospects for assembling larger genomes are discussed.« less

Population-wide sampling of retrotransposon insertion polymorphisms using deep sequencing and efficient detection.

PubMed

Yu, Qichao; Zhang, Wei; Zhang, Xiaolong; Zeng, Yongli; Wang, Yeming; Wang, Yanhui; Xu, Liqin; Huang, Xiaoyun; Li, Nannan; Zhou, Xinlan; Lu, Jie; Guo, Xiaosen; Li, Guibo; Hou, Yong; Liu, Shiping; Li, Bo

2017-09-01

Active retrotransposons play important roles during evolution and continue to shape our genomes today, especially in genetic polymorphisms underlying a diverse set of diseases. However, studies of human retrotransposon insertion polymorphisms (RIPs) based on whole-genome deep sequencing at the population level have not been sufficiently undertaken, despite the obvious need for a thorough characterization of RIPs in the general population. Herein, we present a novel and efficient computational tool called Specific Insertions Detector (SID) for the detection of non-reference RIPs. We demonstrate that SID is suitable for high-depth whole-genome sequencing data using paired-end reads obtained from simulated and real datasets. We construct a comprehensive RIP database using a large population of 90 Han Chinese individuals with a mean ×68 depth per individual. In total, we identify 9342 recent RIPs, and 8433 of these RIPs are novel compared with dbRIP, including 5826 Alu, 2169 long interspersed nuclear element 1 (L1), 383 SVA, and 55 long terminal repeats. Among the 9342 RIPs, 4828 were located in gene regions and 5 were located in protein-coding regions. We demonstrate that RIPs can, in principle, be an informative resource to perform population evolution and phylogenetic analyses. Taking the demographic effects into account, we identify a weak negative selection on SVA and L1 but an approximately neutral selection for Alu elements based on the frequency spectrum of RIPs. SID is a powerful open-source program for the detection of non-reference RIPs. We built a non-reference RIP dataset that greatly enhanced the diversity of RIPs detected in the general population, and it should be invaluable to researchers interested in many aspects of human evolution, genetics, and disease. As a proof of concept, we demonstrate that the RIPs can be used as biomarkers in a similar way as single nucleotide polymorphisms. © The Authors 2017. Published by Oxford University Press.

Benchmarking Deep Learning Models on Large Healthcare Datasets.

PubMed

Purushotham, Sanjay; Meng, Chuizheng; Che, Zhengping; Liu, Yan

2018-06-04

Deep learning models (aka Deep Neural Networks) have revolutionized many fields including computer vision, natural language processing, speech recognition, and is being increasingly used in clinical healthcare applications. However, few works exist which have benchmarked the performance of the deep learning models with respect to the state-of-the-art machine learning models and prognostic scoring systems on publicly available healthcare datasets. In this paper, we present the benchmarking results for several clinical prediction tasks such as mortality prediction, length of stay prediction, and ICD-9 code group prediction using Deep Learning models, ensemble of machine learning models (Super Learner algorithm), SAPS II and SOFA scores. We used the Medical Information Mart for Intensive Care III (MIMIC-III) (v1.4) publicly available dataset, which includes all patients admitted to an ICU at the Beth Israel Deaconess Medical Center from 2001 to 2012, for the benchmarking tasks. Our results show that deep learning models consistently outperform all the other approaches especially when the 'raw' clinical time series data is used as input features to the models. Copyright © 2018 Elsevier Inc. All rights reserved.

CrossQuery: a web tool for easy associative querying of transcriptome data.

PubMed

Wagner, Toni U; Fischer, Andreas; Thoma, Eva C; Schartl, Manfred

2011-01-01

Enormous amounts of data are being generated by modern methods such as transcriptome or exome sequencing and microarray profiling. Primary analyses such as quality control, normalization, statistics and mapping are highly complex and need to be performed by specialists. Thereafter, results are handed back to biomedical researchers, who are then confronted with complicated data lists. For rather simple tasks like data filtering, sorting and cross-association there is a need for new tools which can be used by non-specialists. Here, we describe CrossQuery, a web tool that enables straight forward, simple syntax queries to be executed on transcriptome sequencing and microarray datasets. We provide deep-sequencing data sets of stem cell lines derived from the model fish Medaka and microarray data of human endothelial cells. In the example datasets provided, mRNA expression levels, gene, transcript and sample identification numbers, GO-terms and gene descriptions can be freely correlated, filtered and sorted. Queries can be saved for later reuse and results can be exported to standard formats that allow copy-and-paste to all widespread data visualization tools such as Microsoft Excel. CrossQuery enables researchers to quickly and freely work with transcriptome and microarray data sets requiring only minimal computer skills. Furthermore, CrossQuery allows growing association of multiple datasets as long as at least one common point of correlated information, such as transcript identification numbers or GO-terms, is shared between samples. For advanced users, the object-oriented plug-in and event-driven code design of both server-side and client-side scripts allow easy addition of new features, data sources and data types.

The microbiome of Brazilian mangrove sediments as revealed by metagenomics.

PubMed

Andreote, Fernando Dini; Jiménez, Diego Javier; Chaves, Diego; Dias, Armando Cavalcante Franco; Luvizotto, Danice Mazzer; Dini-Andreote, Francisco; Fasanella, Cristiane Cipola; Lopez, Maryeimy Varon; Baena, Sandra; Taketani, Rodrigo Gouvêa; de Melo, Itamar Soares

2012-01-01

Here we embark in a deep metagenomic survey that revealed the taxonomic and potential metabolic pathways aspects of mangrove sediment microbiology. The extraction of DNA from sediment samples and the direct application of pyrosequencing resulted in approximately 215 Mb of data from four distinct mangrove areas (BrMgv01 to 04) in Brazil. The taxonomic approaches applied revealed the dominance of Deltaproteobacteria and Gammaproteobacteria in the samples. Paired statistical analysis showed higher proportions of specific taxonomic groups in each dataset. The metabolic reconstruction indicated the possible occurrence of processes modulated by the prevailing conditions found in mangrove sediments. In terms of carbon cycling, the sequences indicated the prevalence of genes involved in the metabolism of methane, formaldehyde, and carbon dioxide. With respect to the nitrogen cycle, evidence for sequences associated with dissimilatory reduction of nitrate, nitrogen immobilization, and denitrification was detected. Sequences related to the production of adenylsulfate, sulfite, and H(2)S were relevant to the sulphur cycle. These data indicate that the microbial core involved in methane, nitrogen, and sulphur metabolism consists mainly of Burkholderiaceae, Planctomycetaceae, Rhodobacteraceae, and Desulfobacteraceae. Comparison of our data to datasets from soil and sea samples resulted in the allotment of the mangrove sediments between those samples. The results of this study add valuable data about the composition of microbial communities in mangroves and also shed light on possible transformations promoted by microbial organisms in mangrove sediments.

The Microbiome of Brazilian Mangrove Sediments as Revealed by Metagenomics

PubMed Central

Andreote, Fernando Dini; Jiménez, Diego Javier; Chaves, Diego; Dias, Armando Cavalcante Franco; Luvizotto, Danice Mazzer; Dini-Andreote, Francisco; Fasanella, Cristiane Cipola; Lopez, Maryeimy Varon; Baena, Sandra; Taketani, Rodrigo Gouvêa; de Melo, Itamar Soares

2012-01-01

Here we embark in a deep metagenomic survey that revealed the taxonomic and potential metabolic pathways aspects of mangrove sediment microbiology. The extraction of DNA from sediment samples and the direct application of pyrosequencing resulted in approximately 215 Mb of data from four distinct mangrove areas (BrMgv01 to 04) in Brazil. The taxonomic approaches applied revealed the dominance of Deltaproteobacteria and Gammaproteobacteria in the samples. Paired statistical analysis showed higher proportions of specific taxonomic groups in each dataset. The metabolic reconstruction indicated the possible occurrence of processes modulated by the prevailing conditions found in mangrove sediments. In terms of carbon cycling, the sequences indicated the prevalence of genes involved in the metabolism of methane, formaldehyde, and carbon dioxide. With respect to the nitrogen cycle, evidence for sequences associated with dissimilatory reduction of nitrate, nitrogen immobilization, and denitrification was detected. Sequences related to the production of adenylsulfate, sulfite, and H2S were relevant to the sulphur cycle. These data indicate that the microbial core involved in methane, nitrogen, and sulphur metabolism consists mainly of Burkholderiaceae, Planctomycetaceae, Rhodobacteraceae, and Desulfobacteraceae. Comparison of our data to datasets from soil and sea samples resulted in the allotment of the mangrove sediments between those samples. The results of this study add valuable data about the composition of microbial communities in mangroves and also shed light on possible transformations promoted by microbial organisms in mangrove sediments. PMID:22737213

A deep transcriptomic analysis of pod development in the vanilla orchid (Vanilla planifolia).

PubMed

Rao, Xiaolan; Krom, Nick; Tang, Yuhong; Widiez, Thomas; Havkin-Frenkel, Daphna; Belanger, Faith C; Dixon, Richard A; Chen, Fang

2014-11-07

Pods of the vanilla orchid (Vanilla planifolia) accumulate large amounts of the flavor compound vanillin (3-methoxy, 4-hydroxy-benzaldehyde) as a glucoside during the later stages of their development. At earlier stages, the developing seeds within the pod synthesize a novel lignin polymer, catechyl (C) lignin, in their coats. Genomic resources for determining the biosynthetic routes to these compounds and other flavor components in V. planifolia are currently limited. Using next-generation sequencing technologies, we have generated very large gene sequence datasets from vanilla pods at different times of development, and representing different tissue types, including the seeds, hairs, placental and mesocarp tissues. This developmental series was chosen as being the most informative for interrogation of pathways of vanillin and C-lignin biosynthesis in the pod and seed, respectively. The combined 454/Illumina RNA-seq platforms provide both deep sequence coverage and high quality de novo transcriptome assembly for this non-model crop species. The annotated sequence data provide a foundation for understanding multiple aspects of the biochemistry and development of the vanilla bean, as exemplified by the identification of candidate genes involved in lignin biosynthesis. Our transcriptome data indicate that C-lignin formation in the seed coat involves coordinate expression of monolignol biosynthetic genes with the exception of those encoding the caffeoyl coenzyme A 3-O-methyltransferase for conversion of caffeoyl to feruloyl moieties. This database provides a general resource for further studies on this important flavor species.

Evaluation of Deep Learning Based Stereo Matching Methods: from Ground to Aerial Images

NASA Astrophysics Data System (ADS)

Liu, J.; Ji, S.; Zhang, C.; Qin, Z.

2018-05-01

Dense stereo matching has been extensively studied in photogrammetry and computer vision. In this paper we evaluate the application of deep learning based stereo methods, which were raised from 2016 and rapidly spread, on aerial stereos other than ground images that are commonly used in computer vision community. Two popular methods are evaluated. One learns matching cost with a convolutional neural network (known as MC-CNN); the other produces a disparity map in an end-to-end manner by utilizing both geometry and context (known as GC-net). First, we evaluate the performance of the deep learning based methods for aerial stereo images by a direct model reuse. The models pre-trained on KITTI 2012, KITTI 2015 and Driving datasets separately, are directly applied to three aerial datasets. We also give the results of direct training on target aerial datasets. Second, the deep learning based methods are compared to the classic stereo matching method, Semi-Global Matching(SGM), and a photogrammetric software, SURE, on the same aerial datasets. Third, transfer learning strategy is introduced to aerial image matching based on the assumption of a few target samples available for model fine tuning. It experimentally proved that the conventional methods and the deep learning based methods performed similarly, and the latter had greater potential to be explored.

Discriminative Prediction of A-To-I RNA Editing Events from DNA Sequence

PubMed Central

Sun, Jiangming; Singh, Pratibha; Bagge, Annika; Valtat, Bérengère; Vikman, Petter; Spégel, Peter; Mulder, Hindrik

2016-01-01

RNA editing is a post-transcriptional alteration of RNA sequences that, via insertions, deletions or base substitutions, can affect protein structure as well as RNA and protein expression. Recently, it has been suggested that RNA editing may be more frequent than previously thought. A great impediment, however, to a deeper understanding of this process is the paramount sequencing effort that needs to be undertaken to identify RNA editing events. Here, we describe an in silico approach, based on machine learning, that ameliorates this problem. Using 41 nucleotide long DNA sequences, we show that novel A-to-I RNA editing events can be predicted from known A-to-I RNA editing events intra- and interspecies. The validity of the proposed method was verified in an independent experimental dataset. Using our approach, 203 202 putative A-to-I RNA editing events were predicted in the whole human genome. Out of these, 9% were previously reported. The remaining sites require further validation, e.g., by targeted deep sequencing. In conclusion, the approach described here is a useful tool to identify potential A-to-I RNA editing events without the requirement of extensive RNA sequencing. PMID:27764195

VarDict: a novel and versatile variant caller for next-generation sequencing in cancer research

PubMed Central

Lai, Zhongwu; Markovets, Aleksandra; Ahdesmaki, Miika; Chapman, Brad; Hofmann, Oliver; McEwen, Robert; Johnson, Justin; Dougherty, Brian; Barrett, J. Carl; Dry, Jonathan R.

2016-01-01

Abstract Accurate variant calling in next generation sequencing (NGS) is critical to understand cancer genomes better. Here we present VarDict, a novel and versatile variant caller for both DNA- and RNA-sequencing data. VarDict simultaneously calls SNV, MNV, InDels, complex and structural variants, expanding the detected genetic driver landscape of tumors. It performs local realignments on the fly for more accurate allele frequency estimation. VarDict performance scales linearly to sequencing depth, enabling ultra-deep sequencing used to explore tumor evolution or detect tumor DNA circulating in blood. In addition, VarDict performs amplicon aware variant calling for polymerase chain reaction (PCR)-based targeted sequencing often used in diagnostic settings, and is able to detect PCR artifacts. Finally, VarDict also detects differences in somatic and loss of heterozygosity variants between paired samples. VarDict reprocessing of The Cancer Genome Atlas (TCGA) Lung Adenocarcinoma dataset called known driver mutations in KRAS, EGFR, BRAF, PIK3CA and MET in 16% more patients than previously published variant calls. We believe VarDict will greatly facilitate application of NGS in clinical cancer research. PMID:27060149

Improving Protein Fold Recognition by Deep Learning Networks.

PubMed

Jo, Taeho; Hou, Jie; Eickholt, Jesse; Cheng, Jianlin

2015-12-04

For accurate recognition of protein folds, a deep learning network method (DN-Fold) was developed to predict if a given query-template protein pair belongs to the same structural fold. The input used stemmed from the protein sequence and structural features extracted from the protein pair. We evaluated the performance of DN-Fold along with 18 different methods on Lindahl's benchmark dataset and on a large benchmark set extracted from SCOP 1.75 consisting of about one million protein pairs, at three different levels of fold recognition (i.e., protein family, superfamily, and fold) depending on the evolutionary distance between protein sequences. The correct recognition rate of ensembled DN-Fold for Top 1 predictions is 84.5%, 61.5%, and 33.6% and for Top 5 is 91.2%, 76.5%, and 60.7% at family, superfamily, and fold levels, respectively. We also evaluated the performance of single DN-Fold (DN-FoldS), which showed the comparable results at the level of family and superfamily, compared to ensemble DN-Fold. Finally, we extended the binary classification problem of fold recognition to real-value regression task, which also show a promising performance. DN-Fold is freely available through a web server at http://iris.rnet.missouri.edu/dnfold.

HEp-2 cell image classification method based on very deep convolutional networks with small datasets

NASA Astrophysics Data System (ADS)

Lu, Mengchi; Gao, Long; Guo, Xifeng; Liu, Qiang; Yin, Jianping

2017-07-01

Human Epithelial-2 (HEp-2) cell images staining patterns classification have been widely used to identify autoimmune diseases by the anti-Nuclear antibodies (ANA) test in the Indirect Immunofluorescence (IIF) protocol. Because manual test is time consuming, subjective and labor intensive, image-based Computer Aided Diagnosis (CAD) systems for HEp-2 cell classification are developing. However, methods proposed recently are mostly manual features extraction with low accuracy. Besides, the scale of available benchmark datasets is small, which does not exactly suitable for using deep learning methods. This issue will influence the accuracy of cell classification directly even after data augmentation. To address these issues, this paper presents a high accuracy automatic HEp-2 cell classification method with small datasets, by utilizing very deep convolutional networks (VGGNet). Specifically, the proposed method consists of three main phases, namely image preprocessing, feature extraction and classification. Moreover, an improved VGGNet is presented to address the challenges of small-scale datasets. Experimental results over two benchmark datasets demonstrate that the proposed method achieves superior performance in terms of accuracy compared with existing methods.

Determination of binding affinity upon mutation for type I dockerin-cohesin complexes from Clostridium thermocellum and Clostridium cellulolyticum using deep sequencing.

PubMed

Kowalsky, Caitlin A; Whitehead, Timothy A

2016-12-01

The comprehensive sequence determinants of binding affinity for type I cohesin toward dockerin from Clostridium thermocellum and Clostridium cellulolyticum was evaluated using deep mutational scanning coupled to yeast surface display. We measured the relative binding affinity to dockerin for 2970 and 2778 single point mutants of C. thermocellum and C. cellulolyticum, respectively, representing over 96% of all possible single point mutants. The interface ΔΔG for each variant was reconstructed from sequencing counts and compared with the three independent experimental methods. This reconstruction results in a narrow dynamic range of -0.8-0.5 kcal/mol. The computational software packages FoldX and Rosetta were used to predict mutations that disrupt binding by more than 0.4 kcal/mol. The area under the curve of receiver operator curves was 0.82 for FoldX and 0.77 for Rosetta, showing reasonable agreements between predictions and experimental results. Destabilizing mutations to core and rim positions were predicted with higher accuracy than support positions. This benchmark dataset may be useful for developing new computational prediction tools for the prediction of the mutational effect on binding affinities for protein-protein interactions. Experimental considerations to improve precision and range of the reconstruction method are discussed. Proteins 2016; 84:1914-1928. © 2016 Wiley Periodicals, Inc. © 2016 Wiley Periodicals, Inc.

Deep sequencing of evolving pathogen populations: applications, errors, and bioinformatic solutions

PubMed Central

2014-01-01

Deep sequencing harnesses the high throughput nature of next generation sequencing technologies to generate population samples, treating information contained in individual reads as meaningful. Here, we review applications of deep sequencing to pathogen evolution. Pioneering deep sequencing studies from the virology literature are discussed, such as whole genome Roche-454 sequencing analyses of the dynamics of the rapidly mutating pathogens hepatitis C virus and HIV. Extension of the deep sequencing approach to bacterial populations is then discussed, including the impacts of emerging sequencing technologies. While it is clear that deep sequencing has unprecedented potential for assessing the genetic structure and evolutionary history of pathogen populations, bioinformatic challenges remain. We summarise current approaches to overcoming these challenges, in particular methods for detecting low frequency variants in the context of sequencing error and reconstructing individual haplotypes from short reads. PMID:24428920

Comparing sequencing assays and human-machine analyses in actionable genomics for glioblastoma

PubMed Central

Wrzeszczynski, Kazimierz O.; Frank, Mayu O.; Koyama, Takahiko; Rhrissorrakrai, Kahn; Robine, Nicolas; Utro, Filippo; Emde, Anne-Katrin; Chen, Bo-Juen; Arora, Kanika; Shah, Minita; Vacic, Vladimir; Norel, Raquel; Bilal, Erhan; Bergmann, Ewa A.; Moore Vogel, Julia L.; Bruce, Jeffrey N.; Lassman, Andrew B.; Canoll, Peter; Grommes, Christian; Harvey, Steve; Parida, Laxmi; Michelini, Vanessa V.; Zody, Michael C.; Jobanputra, Vaidehi; Royyuru, Ajay K.

2017-01-01

Objective: To analyze a glioblastoma tumor specimen with 3 different platforms and compare potentially actionable calls from each. Methods: Tumor DNA was analyzed by a commercial targeted panel. In addition, tumor-normal DNA was analyzed by whole-genome sequencing (WGS) and tumor RNA was analyzed by RNA sequencing (RNA-seq). The WGS and RNA-seq data were analyzed by a team of bioinformaticians and cancer oncologists, and separately by IBM Watson Genomic Analytics (WGA), an automated system for prioritizing somatic variants and identifying drugs. Results: More variants were identified by WGS/RNA analysis than by targeted panels. WGA completed a comparable analysis in a fraction of the time required by the human analysts. Conclusions: The development of an effective human-machine interface in the analysis of deep cancer genomic datasets may provide potentially clinically actionable calls for individual patients in a more timely and efficient manner than currently possible. ClinicalTrials.gov identifier: NCT02725684. PMID:28740869

Preparation of metagenomic libraries from naturally occurring marine viruses.

PubMed

Solonenko, Sergei A; Sullivan, Matthew B

2013-01-01

Microbes are now well recognized as major drivers of the biogeochemical cycling that fuels the Earth, and their viruses (phages) are known to be abundant and important in microbial mortality, horizontal gene transfer, and modulating microbial metabolic output. Investigation of environmental phages has been frustrated by an inability to culture the vast majority of naturally occurring diversity coupled with the lack of robust, quantitative, culture-independent methods for studying this uncultured majority. However, for double-stranded DNA phages, a quantitative viral metagenomic sample-to-sequence workflow now exists. Here, we review these advances with special emphasis on the technical details of preparing DNA sequencing libraries for metagenomic sequencing from environmentally relevant low-input DNA samples. Library preparation steps broadly involve manipulating the sample DNA by fragmentation, end repair and adaptor ligation, size fractionation, and amplification. One critical area of future research and development is parallel advances for alternate nucleic acid types such as single-stranded DNA and RNA viruses that are also abundant in nature. Combinations of recent advances in fragmentation (e.g., acoustic shearing and tagmentation), ligation reactions (adaptor-to-template ratio reference table availability), size fractionation (non-gel-sizing), and amplification (linear amplification for deep sequencing and linker amplification protocols) enhance our ability to generate quantitatively representative metagenomic datasets from low-input DNA samples. Such datasets are already providing new insights into the role of viruses in marine systems and will continue to do so as new environments are explored and synergies and paradigms emerge from large-scale comparative analyses. © 2013 Elsevier Inc. All rights reserved.

Estimating time of HIV-1 infection from next-generation sequence diversity

PubMed Central

2017-01-01

Estimating the time since infection (TI) in newly diagnosed HIV-1 patients is challenging, but important to understand the epidemiology of the infection. Here we explore the utility of virus diversity estimated by next-generation sequencing (NGS) as novel biomarker by using a recent genome-wide longitudinal dataset obtained from 11 untreated HIV-1-infected patients with known dates of infection. The results were validated on a second dataset from 31 patients. Virus diversity increased linearly with time, particularly at 3rd codon positions, with little inter-patient variation. The precision of the TI estimate improved with increasing sequencing depth, showing that diversity in NGS data yields superior estimates to the number of ambiguous sites in Sanger sequences, which is one of the alternative biomarkers. The full advantage of deep NGS was utilized with continuous diversity measures such as average pairwise distance or site entropy, rather than the fraction of polymorphic sites. The precision depended on the genomic region and codon position and was highest when 3rd codon positions in the entire pol gene were used. For these data, TI estimates had a mean absolute error of around 1 year. The error increased only slightly from around 0.6 years at a TI of 6 months to around 1.1 years at 6 years. Our results show that virus diversity determined by NGS can be used to estimate time since HIV-1 infection many years after the infection, in contrast to most alternative biomarkers. We provide the regression coefficients as well as web tool for TI estimation. PMID:28968389

Combining deep residual neural network features with supervised machine learning algorithms to classify diverse food image datasets.

PubMed

McAllister, Patrick; Zheng, Huiru; Bond, Raymond; Moorhead, Anne

2018-04-01

Obesity is increasing worldwide and can cause many chronic conditions such as type-2 diabetes, heart disease, sleep apnea, and some cancers. Monitoring dietary intake through food logging is a key method to maintain a healthy lifestyle to prevent and manage obesity. Computer vision methods have been applied to food logging to automate image classification for monitoring dietary intake. In this work we applied pretrained ResNet-152 and GoogleNet convolutional neural networks (CNNs), initially trained using ImageNet Large Scale Visual Recognition Challenge (ILSVRC) dataset with MatConvNet package, to extract features from food image datasets; Food 5K, Food-11, RawFooT-DB, and Food-101. Deep features were extracted from CNNs and used to train machine learning classifiers including artificial neural network (ANN), support vector machine (SVM), Random Forest, and Naive Bayes. Results show that using ResNet-152 deep features with SVM with RBF kernel can accurately detect food items with 99.4% accuracy using Food-5K validation food image dataset and 98.8% with Food-5K evaluation dataset using ANN, SVM-RBF, and Random Forest classifiers. Trained with ResNet-152 features, ANN can achieve 91.34%, 99.28% when applied to Food-11 and RawFooT-DB food image datasets respectively and SVM with RBF kernel can achieve 64.98% with Food-101 image dataset. From this research it is clear that using deep CNN features can be used efficiently for diverse food item image classification. The work presented in this research shows that pretrained ResNet-152 features provide sufficient generalisation power when applied to a range of food image classification tasks. Copyright © 2018 Elsevier Ltd. All rights reserved.

RIEMS: a software pipeline for sensitive and comprehensive taxonomic classification of reads from metagenomics datasets.

PubMed

Scheuch, Matthias; Höper, Dirk; Beer, Martin

2015-03-03

Fuelled by the advent and subsequent development of next generation sequencing technologies, metagenomics became a powerful tool for the analysis of microbial communities both scientifically and diagnostically. The biggest challenge is the extraction of relevant information from the huge sequence datasets generated for metagenomics studies. Although a plethora of tools are available, data analysis is still a bottleneck. To overcome the bottleneck of data analysis, we developed an automated computational workflow called RIEMS - Reliable Information Extraction from Metagenomic Sequence datasets. RIEMS assigns every individual read sequence within a dataset taxonomically by cascading different sequence analyses with decreasing stringency of the assignments using various software applications. After completion of the analyses, the results are summarised in a clearly structured result protocol organised taxonomically. The high accuracy and performance of RIEMS analyses were proven in comparison with other tools for metagenomics data analysis using simulated sequencing read datasets. RIEMS has the potential to fill the gap that still exists with regard to data analysis for metagenomics studies. The usefulness and power of RIEMS for the analysis of genuine sequencing datasets was demonstrated with an early version of RIEMS in 2011 when it was used to detect the orthobunyavirus sequences leading to the discovery of Schmallenberg virus.

DeepMitosis: Mitosis detection via deep detection, verification and segmentation networks.

PubMed

Li, Chao; Wang, Xinggang; Liu, Wenyu; Latecki, Longin Jan

2018-04-01

Mitotic count is a critical predictor of tumor aggressiveness in the breast cancer diagnosis. Nowadays mitosis counting is mainly performed by pathologists manually, which is extremely arduous and time-consuming. In this paper, we propose an accurate method for detecting the mitotic cells from histopathological slides using a novel multi-stage deep learning framework. Our method consists of a deep segmentation network for generating mitosis region when only a weak label is given (i.e., only the centroid pixel of mitosis is annotated), an elaborately designed deep detection network for localizing mitosis by using contextual region information, and a deep verification network for improving detection accuracy by removing false positives. We validate the proposed deep learning method on two widely used Mitosis Detection in Breast Cancer Histological Images (MITOSIS) datasets. Experimental results show that we can achieve the highest F-score on the MITOSIS dataset from ICPR 2012 grand challenge merely using the deep detection network. For the ICPR 2014 MITOSIS dataset that only provides the centroid location of mitosis, we employ the segmentation model to estimate the bounding box annotation for training the deep detection network. We also apply the verification model to eliminate some false positives produced from the detection model. By fusing scores of the detection and verification models, we achieve the state-of-the-art results. Moreover, our method is very fast with GPU computing, which makes it feasible for clinical practice. Copyright © 2018 Elsevier B.V. All rights reserved.

Benchmark Dataset for Whole Genome Sequence Compression.

PubMed

C L, Biji; S Nair, Achuthsankar

2017-01-01

The research in DNA data compression lacks a standard dataset to test out compression tools specific to DNA. This paper argues that the current state of achievement in DNA compression is unable to be benchmarked in the absence of such scientifically compiled whole genome sequence dataset and proposes a benchmark dataset using multistage sampling procedure. Considering the genome sequence of organisms available in the National Centre for Biotechnology and Information (NCBI) as the universe, the proposed dataset selects 1,105 prokaryotes, 200 plasmids, 164 viruses, and 65 eukaryotes. This paper reports the results of using three established tools on the newly compiled dataset and show that their strength and weakness are evident only with a comparison based on the scientifically compiled benchmark dataset. The sample dataset and the respective links are available @ https://sourceforge.net/projects/benchmarkdnacompressiondataset/.

An improved advertising CTR prediction approach based on the fuzzy deep neural network

PubMed Central

Gao, Shu; Li, Mingjiang

2018-01-01

Combining a deep neural network with fuzzy theory, this paper proposes an advertising click-through rate (CTR) prediction approach based on a fuzzy deep neural network (FDNN). In this approach, fuzzy Gaussian-Bernoulli restricted Boltzmann machine (FGBRBM) is first applied to input raw data from advertising datasets. Next, fuzzy restricted Boltzmann machine (FRBM) is used to construct the fuzzy deep belief network (FDBN) with the unsupervised method layer by layer. Finally, fuzzy logistic regression (FLR) is utilized for modeling the CTR. The experimental results show that the proposed FDNN model outperforms several baseline models in terms of both data representation capability and robustness in advertising click log datasets with noise. PMID:29727443

An improved advertising CTR prediction approach based on the fuzzy deep neural network.

PubMed

Jiang, Zilong; Gao, Shu; Li, Mingjiang

2018-01-01

Combining a deep neural network with fuzzy theory, this paper proposes an advertising click-through rate (CTR) prediction approach based on a fuzzy deep neural network (FDNN). In this approach, fuzzy Gaussian-Bernoulli restricted Boltzmann machine (FGBRBM) is first applied to input raw data from advertising datasets. Next, fuzzy restricted Boltzmann machine (FRBM) is used to construct the fuzzy deep belief network (FDBN) with the unsupervised method layer by layer. Finally, fuzzy logistic regression (FLR) is utilized for modeling the CTR. The experimental results show that the proposed FDNN model outperforms several baseline models in terms of both data representation capability and robustness in advertising click log datasets with noise.

A high-throughput system for high-quality tomographic reconstruction of large datasets at Diamond Light Source

PubMed Central

Atwood, Robert C.; Bodey, Andrew J.; Price, Stephen W. T.; Basham, Mark; Drakopoulos, Michael

2015-01-01

Tomographic datasets collected at synchrotrons are becoming very large and complex, and, therefore, need to be managed efficiently. Raw images may have high pixel counts, and each pixel can be multidimensional and associated with additional data such as those derived from spectroscopy. In time-resolved studies, hundreds of tomographic datasets can be collected in sequence, yielding terabytes of data. Users of tomographic beamlines are drawn from various scientific disciplines, and many are keen to use tomographic reconstruction software that does not require a deep understanding of reconstruction principles. We have developed Savu, a reconstruction pipeline that enables users to rapidly reconstruct data to consistently create high-quality results. Savu is designed to work in an ‘orthogonal’ fashion, meaning that data can be converted between projection and sinogram space throughout the processing workflow as required. The Savu pipeline is modular and allows processing strategies to be optimized for users' purposes. In addition to the reconstruction algorithms themselves, it can include modules for identification of experimental problems, artefact correction, general image processing and data quality assessment. Savu is open source, open licensed and ‘facility-independent’: it can run on standard cluster infrastructure at any institution. PMID:25939626

Deep learning and face recognition: the state of the art

NASA Astrophysics Data System (ADS)

Balaban, Stephen

2015-05-01

Deep Neural Networks (DNNs) have established themselves as a dominant technique in machine learning. DNNs have been top performers on a wide variety of tasks including image classification, speech recognition, and face recognition.1-3 Convolutional neural networks (CNNs) have been used in nearly all of the top performing methods on the Labeled Faces in the Wild (LFW) dataset.3-6 In this talk and accompanying paper, I attempt to provide a review and summary of the deep learning techniques used in the state-of-the-art. In addition, I highlight the need for both larger and more challenging public datasets to benchmark these systems. Despite the ability of DNNs and autoencoders to perform unsupervised feature learning, modern facial recognition pipelines still require domain specific engineering in the form of re-alignment. For example, in Facebook's recent DeepFace paper, a 3D "frontalization" step lies at the beginning of the pipeline. This step creates a 3D face model for the incoming image and then uses a series of affine transformations of the fiducial points to "frontalize" the image. This step enables the DeepFace system to use a neural network architecture with locally connected layers without weight sharing as opposed to standard convolutional layers.6 Deep learning techniques combined with large datasets have allowed research groups to surpass human level performance on the LFW dataset.3, 5 The high accuracy (99.63% for FaceNet at the time of publishing) and utilization of outside data (hundreds of millions of images in the case of Google's FaceNet) suggest that current face verification benchmarks such as LFW may not be challenging enough, nor provide enough data, for current techniques.3, 5 There exist a variety of organizations with mobile photo sharing applications that would be capable of releasing a very large scale and highly diverse dataset of facial images captured on mobile devices. Such an "ImageNet for Face Recognition" would likely receive a warm welcome from researchers and practitioners alike.

Improving resolution of MR images with an adversarial network incorporating images with different contrast.

PubMed

Kim, Ki Hwan; Do, Won-Joon; Park, Sung-Hong

2018-05-04

The routine MRI scan protocol consists of multiple pulse sequences that acquire images of varying contrast. Since high frequency contents such as edges are not significantly affected by image contrast, down-sampled images in one contrast may be improved by high resolution (HR) images acquired in another contrast, reducing the total scan time. In this study, we propose a new deep learning framework that uses HR MR images in one contrast to generate HR MR images from highly down-sampled MR images in another contrast. The proposed convolutional neural network (CNN) framework consists of two CNNs: (a) a reconstruction CNN for generating HR images from the down-sampled images using HR images acquired with a different MRI sequence and (b) a discriminator CNN for improving the perceptual quality of the generated HR images. The proposed method was evaluated using a public brain tumor database and in vivo datasets. The performance of the proposed method was assessed in tumor and no-tumor cases separately, with perceptual image quality being judged by a radiologist. To overcome the challenge of training the network with a small number of available in vivo datasets, the network was pretrained using the public database and then fine-tuned using the small number of in vivo datasets. The performance of the proposed method was also compared to that of several compressed sensing (CS) algorithms. Incorporating HR images of another contrast improved the quantitative assessments of the generated HR image in reference to ground truth. Also, incorporating a discriminator CNN yielded perceptually higher image quality. These results were verified in regions of normal tissue as well as tumors for various MRI sequences from pseudo k-space data generated from the public database. The combination of pretraining with the public database and fine-tuning with the small number of real k-space datasets enhanced the performance of CNNs in in vivo application compared to training CNNs from scratch. The proposed method outperformed the compressed sensing methods. The proposed method can be a good strategy for accelerating routine MRI scanning. © 2018 American Association of Physicists in Medicine.

A novel on-line spatial-temporal k-anonymity method for location privacy protection from sequence rules-based inference attacks.

PubMed

Zhang, Haitao; Wu, Chenxue; Chen, Zewei; Liu, Zhao; Zhu, Yunhong

2017-01-01

Analyzing large-scale spatial-temporal k-anonymity datasets recorded in location-based service (LBS) application servers can benefit some LBS applications. However, such analyses can allow adversaries to make inference attacks that cannot be handled by spatial-temporal k-anonymity methods or other methods for protecting sensitive knowledge. In response to this challenge, first we defined a destination location prediction attack model based on privacy-sensitive sequence rules mined from large scale anonymity datasets. Then we proposed a novel on-line spatial-temporal k-anonymity method that can resist such inference attacks. Our anti-attack technique generates new anonymity datasets with awareness of privacy-sensitive sequence rules. The new datasets extend the original sequence database of anonymity datasets to hide the privacy-sensitive rules progressively. The process includes two phases: off-line analysis and on-line application. In the off-line phase, sequence rules are mined from an original sequence database of anonymity datasets, and privacy-sensitive sequence rules are developed by correlating privacy-sensitive spatial regions with spatial grid cells among the sequence rules. In the on-line phase, new anonymity datasets are generated upon LBS requests by adopting specific generalization and avoidance principles to hide the privacy-sensitive sequence rules progressively from the extended sequence anonymity datasets database. We conducted extensive experiments to test the performance of the proposed method, and to explore the influence of the parameter K value. The results demonstrated that our proposed approach is faster and more effective for hiding privacy-sensitive sequence rules in terms of hiding sensitive rules ratios to eliminate inference attacks. Our method also had fewer side effects in terms of generating new sensitive rules ratios than the traditional spatial-temporal k-anonymity method, and had basically the same side effects in terms of non-sensitive rules variation ratios with the traditional spatial-temporal k-anonymity method. Furthermore, we also found the performance variation tendency from the parameter K value, which can help achieve the goal of hiding the maximum number of original sensitive rules while generating a minimum of new sensitive rules and affecting a minimum number of non-sensitive rules.

A novel on-line spatial-temporal k-anonymity method for location privacy protection from sequence rules-based inference attacks

PubMed Central

Wu, Chenxue; Liu, Zhao; Zhu, Yunhong

2017-01-01

Analyzing large-scale spatial-temporal k-anonymity datasets recorded in location-based service (LBS) application servers can benefit some LBS applications. However, such analyses can allow adversaries to make inference attacks that cannot be handled by spatial-temporal k-anonymity methods or other methods for protecting sensitive knowledge. In response to this challenge, first we defined a destination location prediction attack model based on privacy-sensitive sequence rules mined from large scale anonymity datasets. Then we proposed a novel on-line spatial-temporal k-anonymity method that can resist such inference attacks. Our anti-attack technique generates new anonymity datasets with awareness of privacy-sensitive sequence rules. The new datasets extend the original sequence database of anonymity datasets to hide the privacy-sensitive rules progressively. The process includes two phases: off-line analysis and on-line application. In the off-line phase, sequence rules are mined from an original sequence database of anonymity datasets, and privacy-sensitive sequence rules are developed by correlating privacy-sensitive spatial regions with spatial grid cells among the sequence rules. In the on-line phase, new anonymity datasets are generated upon LBS requests by adopting specific generalization and avoidance principles to hide the privacy-sensitive sequence rules progressively from the extended sequence anonymity datasets database. We conducted extensive experiments to test the performance of the proposed method, and to explore the influence of the parameter K value. The results demonstrated that our proposed approach is faster and more effective for hiding privacy-sensitive sequence rules in terms of hiding sensitive rules ratios to eliminate inference attacks. Our method also had fewer side effects in terms of generating new sensitive rules ratios than the traditional spatial-temporal k-anonymity method, and had basically the same side effects in terms of non-sensitive rules variation ratios with the traditional spatial-temporal k-anonymity method. Furthermore, we also found the performance variation tendency from the parameter K value, which can help achieve the goal of hiding the maximum number of original sensitive rules while generating a minimum of new sensitive rules and affecting a minimum number of non-sensitive rules. PMID:28767687

Genome-wide assessment of differential translations with ribosome profiling data

PubMed Central

Xiao, Zhengtao; Zou, Qin; Liu, Yu; Yang, Xuerui

2016-01-01

The closely regulated process of mRNA translation is crucial for precise control of protein abundance and quality. Ribosome profiling, a combination of ribosome foot-printing and RNA deep sequencing, has been used in a large variety of studies to quantify genome-wide mRNA translation. Here, we developed Xtail, an analysis pipeline tailored for ribosome profiling data that comprehensively and accurately identifies differentially translated genes in pairwise comparisons. Applied on simulated and real datasets, Xtail exhibits high sensitivity with minimal false-positive rates, outperforming existing methods in the accuracy of quantifying differential translations. With published ribosome profiling datasets, Xtail does not only reveal differentially translated genes that make biological sense, but also uncovers new events of differential translation in human cancer cells on mTOR signalling perturbation and in human primary macrophages on interferon gamma (IFN-γ) treatment. This demonstrates the value of Xtail in providing novel insights into the molecular mechanisms that involve translational dysregulations. PMID:27041671

Clustering Single-Cell Expression Data Using Random Forest Graphs.

PubMed

Pouyan, Maziyar Baran; Nourani, Mehrdad

2017-07-01

Complex tissues such as brain and bone marrow are made up of multiple cell types. As the study of biological tissue structure progresses, the role of cell-type-specific research becomes increasingly important. Novel sequencing technology such as single-cell cytometry provides researchers access to valuable biological data. Applying machine-learning techniques to these high-throughput datasets provides deep insights into the cellular landscape of the tissue where those cells are a part of. In this paper, we propose the use of random-forest-based single-cell profiling, a new machine-learning-based technique, to profile different cell types of intricate tissues using single-cell cytometry data. Our technique utilizes random forests to capture cell marker dependences and model the cellular populations using the cell network concept. This cellular network helps us discover what cell types are in the tissue. Our experimental results on public-domain datasets indicate promising performance and accuracy of our technique in extracting cell populations of complex tissues.

Pubface: Celebrity face identification based on deep learning

NASA Astrophysics Data System (ADS)

Ouanan, H.; Ouanan, M.; Aksasse, B.

2018-05-01

In this paper, we describe a new real time application called PubFace, which allows to recognize celebrities in public spaces by employs a new pose invariant face recognition deep neural network algorithm with an extremely low error rate. To build this application, we make the following contributions: firstly, we build a novel dataset with over five million faces labelled. Secondly, we fine tuning the deep convolutional neural network (CNN) VGG-16 architecture on our new dataset that we have built. Finally, we deploy this model on the Raspberry Pi 3 model B using the OpenCv dnn module (OpenCV 3.3).

Prediction of enhancer-promoter interactions via natural language processing.

PubMed

Zeng, Wanwen; Wu, Mengmeng; Jiang, Rui

2018-05-09

Precise identification of three-dimensional genome organization, especially enhancer-promoter interactions (EPIs), is important to deciphering gene regulation, cell differentiation and disease mechanisms. Currently, it is a challenging task to distinguish true interactions from other nearby non-interacting ones since the power of traditional experimental methods is limited due to low resolution or low throughput. We propose a novel computational framework EP2vec to assay three-dimensional genomic interactions. We first extract sequence embedding features, defined as fixed-length vector representations learned from variable-length sequences using an unsupervised deep learning method in natural language processing. Then, we train a classifier to predict EPIs using the learned representations in supervised way. Experimental results demonstrate that EP2vec obtains F1 scores ranging from 0.841~ 0.933 on different datasets, which outperforms existing methods. We prove the robustness of sequence embedding features by carrying out sensitivity analysis. Besides, we identify motifs that represent cell line-specific information through analysis of the learned sequence embedding features by adopting attention mechanism. Last, we show that even superior performance with F1 scores 0.889~ 0.940 can be achieved by combining sequence embedding features and experimental features. EP2vec sheds light on feature extraction for DNA sequences of arbitrary lengths and provides a powerful approach for EPIs identification.

Genomics dataset of unidentified disclosed isolates.

PubMed

Rekadwad, Bhagwan N

2016-09-01

Analysis of DNA sequences is necessary for higher hierarchical classification of the organisms. It gives clues about the characteristics of organisms and their taxonomic position. This dataset is chosen to find complexities in the unidentified DNA in the disclosed patents. A total of 17 unidentified DNA sequences were thoroughly analyzed. The quick response codes were generated. AT/GC content of the DNA sequences analysis was carried out. The QR is helpful for quick identification of isolates. AT/GC content is helpful for studying their stability at different temperatures. Additionally, a dataset on cleavage code and enzyme code studied under the restriction digestion study, which helpful for performing studies using short DNA sequences was reported. The dataset disclosed here is the new revelatory data for exploration of unique DNA sequences for evaluation, identification, comparison and analysis.

Automated analysis of high-content microscopy data with deep learning.

PubMed

Kraus, Oren Z; Grys, Ben T; Ba, Jimmy; Chong, Yolanda; Frey, Brendan J; Boone, Charles; Andrews, Brenda J

2017-04-18

Existing computational pipelines for quantitative analysis of high-content microscopy data rely on traditional machine learning approaches that fail to accurately classify more than a single dataset without substantial tuning and training, requiring extensive analysis. Here, we demonstrate that the application of deep learning to biological image data can overcome the pitfalls associated with conventional machine learning classifiers. Using a deep convolutional neural network (DeepLoc) to analyze yeast cell images, we show improved performance over traditional approaches in the automated classification of protein subcellular localization. We also demonstrate the ability of DeepLoc to classify highly divergent image sets, including images of pheromone-arrested cells with abnormal cellular morphology, as well as images generated in different genetic backgrounds and in different laboratories. We offer an open-source implementation that enables updating DeepLoc on new microscopy datasets. This study highlights deep learning as an important tool for the expedited analysis of high-content microscopy data. © 2017 The Authors. Published under the terms of the CC BY 4.0 license.

De novo Transcriptome Analysis of Portunus trituberculatus Ovary and Testis by RNA-Seq: Identification of Genes Involved in Gonadal Development

PubMed Central

Meng, Xian-liang; Liu, Ping; Jia, Fu-long; Li, Jian; Gao, Bao-Quan

2015-01-01

The swimming crab Portunus trituberculatus is a commercially important crab species in East Asia countries. Gonadal development is a physiological process of great significance to the reproduction as well as commercial seed production for P. trituberculatus. However, little is currently known about the molecular mechanisms governing the developmental processes of gonads in this species. To open avenues of molecular research on P. trituberculatus gonadal development, Illumina paired-end sequencing technology was employed to develop deep-coverage transcriptome sequencing data for its gonads. Illumina sequencing generated 58,429,148 and 70,474,978 high-quality reads from the ovary and testis cDNA library, respectively. All these reads were assembled into 54,960 unigenes with an average sequence length of 879 bp, of which 12,340 unigenes (22.45% of the total) matched sequences in GenBank non-redundant database. Based on our transcriptome analysis as well as published literature, a number of candidate genes potentially involved in the regulation of gonadal development of P. trituberculatus were identified, such as FAOMeT, mPRγ, PGMRC1, PGDS, PGER4, 3β-HSD and 17β-HSDs. Differential expression analysis generated 5,919 differentially expressed genes between ovary and testis, among which many genes related to gametogenesis and several genes previously reported to be critical in differentiation and development of gonads were found, including Foxl2, Wnt4, Fst, Fem-1 and Sox9. Furthermore, 28,534 SSRs and 111,646 high-quality SNPs were identified in this transcriptome dataset. This work represents the first transcriptome analysis of P. trituberculatus gonads using the next generation sequencing technology and provides a valuable dataset for understanding molecular mechanisms controlling development of gonads and facilitating future investigation of reproductive biology in this species. The molecular markers obtained in this study will provide a fundamental basis for population genetics and functional genomics in P. trituberculatus and other closely related species. PMID:26042806

DeepBase: annotation and discovery of microRNAs and other noncoding RNAs from deep-sequencing data.

PubMed

Yang, Jian-Hua; Qu, Liang-Hu

2012-01-01

Recent advances in high-throughput deep-sequencing technology have produced large numbers of short and long RNA sequences and enabled the detection and profiling of known and novel microRNAs (miRNAs) and other noncoding RNAs (ncRNAs) at unprecedented sensitivity and depth. In this chapter, we describe the use of deepBase, a database that we have developed to integrate all public deep-sequencing data and to facilitate the comprehensive annotation and discovery of miRNAs and other ncRNAs from these data. deepBase provides an integrative, interactive, and versatile web graphical interface to evaluate miRBase-annotated miRNA genes and other known ncRNAs, explores the expression patterns of miRNAs and other ncRNAs, and discovers novel miRNAs and other ncRNAs from deep-sequencing data. deepBase also provides a deepView genome browser to comparatively analyze these data at multiple levels. deepBase is available at http://deepbase.sysu.edu.cn/.

Detection and sequence/structure mapping of biophysical constraints to protein variation in saturated mutational libraries and protein sequence alignments with a dedicated server.

PubMed

Abriata, Luciano A; Bovigny, Christophe; Dal Peraro, Matteo

2016-06-17

Protein variability can now be studied by measuring high-resolution tolerance-to-substitution maps and fitness landscapes in saturated mutational libraries. But these rich and expensive datasets are typically interpreted coarsely, restricting detailed analyses to positions of extremely high or low variability or dubbed important beforehand based on existing knowledge about active sites, interaction surfaces, (de)stabilizing mutations, etc. Our new webserver PsychoProt (freely available without registration at http://psychoprot.epfl.ch or at http://lucianoabriata.altervista.org/psychoprot/index.html ) helps to detect, quantify, and sequence/structure map the biophysical and biochemical traits that shape amino acid preferences throughout a protein as determined by deep-sequencing of saturated mutational libraries or from large alignments of naturally occurring variants. We exemplify how PsychoProt helps to (i) unveil protein structure-function relationships from experiments and from alignments that are consistent with structures according to coevolution analysis, (ii) recall global information about structural and functional features and identify hitherto unknown constraints to variation in alignments, and (iii) point at different sources of variation among related experimental datasets or between experimental and alignment-based data. Remarkably, metabolic costs of the amino acids pose strong constraints to variability at protein surfaces in nature but not in the laboratory. This and other differences call for caution when extrapolating results from in vitro experiments to natural scenarios in, for example, studies of protein evolution. We show through examples how PsychoProt can be a useful tool for the broad communities of structural biology and molecular evolution, particularly for studies about protein modeling, evolution and design.

Making sense of deep sequencing

PubMed Central

Goldman, D.; Domschke, K.

2016-01-01

This review, the first of an occasional series, tries to make sense of the concepts and uses of deep sequencing of polynucleic acids (DNA and RNA). Deep sequencing, synonymous with next-generation sequencing, high-throughput sequencing and massively parallel sequencing, includes whole genome sequencing but is more often and diversely applied to specific parts of the genome captured in different ways, for example the highly expressed portion of the genome known as the exome and portions of the genome that are epigenetically marked either by DNA methylation, the binding of proteins including histones, or that are in different configurations and thus more or less accessible to enzymes that cleave DNA. Deep sequencing of RNA (RNASeq) reverse-transcribed to complementary DNA is invaluable for measuring RNA expression and detecting changes in RNA structure. Important concepts in deep sequencing include the length and depth of sequence reads, mapping and assembly of reads, sequencing error, haplotypes, and the propensity of deep sequencing, as with other types of ‘big data’, to generate large numbers of errors, requiring monitoring for methodologic biases and strategies for replication and validation. Deep sequencing yields a unique genetic fingerprint that can be used to identify a person, and a trove of predictors of genetic medical diseases. Deep sequencing to identify epigenetic events including changes in DNA methylation and RNA expression can reveal the history and impact of environmental exposures. Because of the power of sequencing to identify and deliver biomedically significant information about a person and their blood relatives, it creates ethical dilemmas and practical challenges in research and clinical care, for example the decision and procedures to report incidental findings that will increasingly and frequently be discovered. PMID:24925306

Improving Protein Fold Recognition by Deep Learning Networks

NASA Astrophysics Data System (ADS)

Jo, Taeho; Hou, Jie; Eickholt, Jesse; Cheng, Jianlin

2015-12-01

For accurate recognition of protein folds, a deep learning network method (DN-Fold) was developed to predict if a given query-template protein pair belongs to the same structural fold. The input used stemmed from the protein sequence and structural features extracted from the protein pair. We evaluated the performance of DN-Fold along with 18 different methods on Lindahl’s benchmark dataset and on a large benchmark set extracted from SCOP 1.75 consisting of about one million protein pairs, at three different levels of fold recognition (i.e., protein family, superfamily, and fold) depending on the evolutionary distance between protein sequences. The correct recognition rate of ensembled DN-Fold for Top 1 predictions is 84.5%, 61.5%, and 33.6% and for Top 5 is 91.2%, 76.5%, and 60.7% at family, superfamily, and fold levels, respectively. We also evaluated the performance of single DN-Fold (DN-FoldS), which showed the comparable results at the level of family and superfamily, compared to ensemble DN-Fold. Finally, we extended the binary classification problem of fold recognition to real-value regression task, which also show a promising performance. DN-Fold is freely available through a web server at http://iris.rnet.missouri.edu/dnfold.

A reference human genome dataset of the BGISEQ-500 sequencer.

PubMed

Huang, Jie; Liang, Xinming; Xuan, Yuankai; Geng, Chunyu; Li, Yuxiang; Lu, Haorong; Qu, Shoufang; Mei, Xianglin; Chen, Hongbo; Yu, Ting; Sun, Nan; Rao, Junhua; Wang, Jiahao; Zhang, Wenwei; Chen, Ying; Liao, Sha; Jiang, Hui; Liu, Xin; Yang, Zhaopeng; Mu, Feng; Gao, Shangxian

2017-05-01

BGISEQ-500 is a new desktop sequencer developed by BGI. Using DNA nanoball and combinational probe anchor synthesis developed from Complete Genomics™ sequencing technologies, it generates short reads at a large scale. Here, we present the first human whole-genome sequencing dataset of BGISEQ-500. The dataset was generated by sequencing the widely used cell line HG001 (NA12878) in two sequencing runs of paired-end 50 bp (PE50) and two sequencing runs of paired-end 100 bp (PE100). We also include examples of the raw images from the sequencer for reference. Finally, we identified variations using this dataset, estimated the accuracy of the variations, and compared to that of the variations identified from similar amounts of publicly available HiSeq2500 data. We found similar single nucleotide polymorphism (SNP) detection accuracy for the BGISEQ-500 PE100 data (false positive rate [FPR] = 0.00020%, sensitivity = 96.20%) compared to the PE150 HiSeq2500 data (FPR = 0.00017%, sensitivity = 96.60%) better SNP detection accuracy than the PE50 data (FPR = 0.0006%, sensitivity = 94.15%). But for insertions and deletions (indels), we found lower accuracy for BGISEQ-500 data (FPR = 0.00069% and 0.00067% for PE100 and PE50 respectively, sensitivity = 88.52% and 70.93%) than the HiSeq2500 data (FPR = 0.00032%, sensitivity = 96.28%). Our dataset can serve as the reference dataset, providing basic information not just for future development, but also for all research and applications based on the new sequencing platform. © The Authors 2017. Published by Oxford University Press.

P-Hint-Hunt: a deep parallelized whole genome DNA methylation detection tool.

PubMed

Peng, Shaoliang; Yang, Shunyun; Gao, Ming; Liao, Xiangke; Liu, Jie; Yang, Canqun; Wu, Chengkun; Yu, Wenqiang

2017-03-14

The increasing studies have been conducted using whole genome DNA methylation detection as one of the most important part of epigenetics research to find the significant relationships among DNA methylation and several typical diseases, such as cancers and diabetes. In many of those studies, mapping the bisulfite treated sequence to the whole genome has been the main method to study DNA cytosine methylation. However, today's relative tools almost suffer from inaccuracies and time-consuming problems. In our study, we designed a new DNA methylation prediction tool ("Hint-Hunt") to solve the problem. By having an optimal complex alignment computation and Smith-Waterman matrix dynamic programming, Hint-Hunt could analyze and predict the DNA methylation status. But when Hint-Hunt tried to predict DNA methylation status with large-scale dataset, there are still slow speed and low temporal-spatial efficiency problems. In order to solve the problems of Smith-Waterman dynamic programming and low temporal-spatial efficiency, we further design a deep parallelized whole genome DNA methylation detection tool ("P-Hint-Hunt") on Tianhe-2 (TH-2) supercomputer. To the best of our knowledge, P-Hint-Hunt is the first parallel DNA methylation detection tool with a high speed-up to process large-scale dataset, and could run both on CPU and Intel Xeon Phi coprocessors. Moreover, we deploy and evaluate Hint-Hunt and P-Hint-Hunt on TH-2 supercomputer in different scales. The experimental results illuminate our tools eliminate the deviation caused by bisulfite treatment in mapping procedure and the multi-level parallel program yields a 48 times speed-up with 64 threads. P-Hint-Hunt gain a deep acceleration on CPU and Intel Xeon Phi heterogeneous platform, which gives full play of the advantages of multi-cores (CPU) and many-cores (Phi).

Computational optical tomography using 3-D deep convolutional neural networks

NASA Astrophysics Data System (ADS)

Nguyen, Thanh; Bui, Vy; Nehmetallah, George

2018-04-01

Deep convolutional neural networks (DCNNs) offer a promising performance for many image processing areas, such as super-resolution, deconvolution, image classification, denoising, and segmentation, with outstanding results. Here, we develop for the first time, to our knowledge, a method to perform 3-D computational optical tomography using 3-D DCNN. A simulated 3-D phantom dataset was first constructed and converted to a dataset of phase objects imaged on a spatial light modulator. For each phase image in the dataset, the corresponding diffracted intensity image was experimentally recorded on a CCD. We then experimentally demonstrate the ability of the developed 3-D DCNN algorithm to solve the inverse problem by reconstructing the 3-D index of refraction distributions of test phantoms from the dataset from their corresponding diffraction patterns.

DAMe: a toolkit for the initial processing of datasets with PCR replicates of double-tagged amplicons for DNA metabarcoding analyses.

PubMed

Zepeda-Mendoza, Marie Lisandra; Bohmann, Kristine; Carmona Baez, Aldo; Gilbert, M Thomas P

2016-05-03

DNA metabarcoding is an approach for identifying multiple taxa in an environmental sample using specific genetic loci and taxa-specific primers. When combined with high-throughput sequencing it enables the taxonomic characterization of large numbers of samples in a relatively time- and cost-efficient manner. One recent laboratory development is the addition of 5'-nucleotide tags to both primers producing double-tagged amplicons and the use of multiple PCR replicates to filter erroneous sequences. However, there is currently no available toolkit for the straightforward analysis of datasets produced in this way. We present DAMe, a toolkit for the processing of datasets generated by double-tagged amplicons from multiple PCR replicates derived from an unlimited number of samples. Specifically, DAMe can be used to (i) sort amplicons by tag combination, (ii) evaluate PCR replicates dissimilarity, and (iii) filter sequences derived from sequencing/PCR errors, chimeras, and contamination. This is attained by calculating the following parameters: (i) sequence content similarity between the PCR replicates from each sample, (ii) reproducibility of each unique sequence across the PCR replicates, and (iii) copy number of the unique sequences in each PCR replicate. We showcase the insights that can be obtained using DAMe prior to taxonomic assignment, by applying it to two real datasets that vary in their complexity regarding number of samples, sequencing libraries, PCR replicates, and used tag combinations. Finally, we use a third mock dataset to demonstrate the impact and importance of filtering the sequences with DAMe. DAMe allows the user-friendly manipulation of amplicons derived from multiple samples with PCR replicates built in a single or multiple sequencing libraries. It allows the user to: (i) collapse amplicons into unique sequences and sort them by tag combination while retaining the sample identifier and copy number information, (ii) identify sequences carrying unused tag combinations, (iii) evaluate the comparability of PCR replicates of the same sample, and (iv) filter tagged amplicons from a number of PCR replicates using parameters of minimum length, copy number, and reproducibility across the PCR replicates. This enables an efficient analysis of complex datasets, and ultimately increases the ease of handling datasets from large-scale studies.

Quantum-assisted Helmholtz machines: A quantum–classical deep learning framework for industrial datasets in near-term devices

NASA Astrophysics Data System (ADS)

Benedetti, Marcello; Realpe-Gómez, John; Perdomo-Ortiz, Alejandro

2018-07-01

Machine learning has been presented as one of the key applications for near-term quantum technologies, given its high commercial value and wide range of applicability. In this work, we introduce the quantum-assisted Helmholtz machine:a hybrid quantum–classical framework with the potential of tackling high-dimensional real-world machine learning datasets on continuous variables. Instead of using quantum computers only to assist deep learning, as previous approaches have suggested, we use deep learning to extract a low-dimensional binary representation of data, suitable for processing on relatively small quantum computers. Then, the quantum hardware and deep learning architecture work together to train an unsupervised generative model. We demonstrate this concept using 1644 quantum bits of a D-Wave 2000Q quantum device to model a sub-sampled version of the MNIST handwritten digit dataset with 16 × 16 continuous valued pixels. Although we illustrate this concept on a quantum annealer, adaptations to other quantum platforms, such as ion-trap technologies or superconducting gate-model architectures, could be explored within this flexible framework.

De novo peptide sequencing by deep learning

PubMed Central

Tran, Ngoc Hieu; Zhang, Xianglilan; Xin, Lei; Shan, Baozhen; Li, Ming

2017-01-01

De novo peptide sequencing from tandem MS data is the key technology in proteomics for the characterization of proteins, especially for new sequences, such as mAbs. In this study, we propose a deep neural network model, DeepNovo, for de novo peptide sequencing. DeepNovo architecture combines recent advances in convolutional neural networks and recurrent neural networks to learn features of tandem mass spectra, fragment ions, and sequence patterns of peptides. The networks are further integrated with local dynamic programming to solve the complex optimization task of de novo sequencing. We evaluated the method on a wide variety of species and found that DeepNovo considerably outperformed state of the art methods, achieving 7.7–22.9% higher accuracy at the amino acid level and 38.1–64.0% higher accuracy at the peptide level. We further used DeepNovo to automatically reconstruct the complete sequences of antibody light and heavy chains of mouse, achieving 97.5–100% coverage and 97.2–99.5% accuracy, without assisting databases. Moreover, DeepNovo is retrainable to adapt to any sources of data and provides a complete end-to-end training and prediction solution to the de novo sequencing problem. Not only does our study extend the deep learning revolution to a new field, but it also shows an innovative approach in solving optimization problems by using deep learning and dynamic programming. PMID:28720701

White blood cells identification system based on convolutional deep neural learning networks.

PubMed

Shahin, A I; Guo, Yanhui; Amin, K M; Sharawi, Amr A

2017-11-16

White blood cells (WBCs) differential counting yields valued information about human health and disease. The current developed automated cell morphology equipments perform differential count which is based on blood smear image analysis. Previous identification systems for WBCs consist of successive dependent stages; pre-processing, segmentation, feature extraction, feature selection, and classification. There is a real need to employ deep learning methodologies so that the performance of previous WBCs identification systems can be increased. Classifying small limited datasets through deep learning systems is a major challenge and should be investigated. In this paper, we propose a novel identification system for WBCs based on deep convolutional neural networks. Two methodologies based on transfer learning are followed: transfer learning based on deep activation features and fine-tuning of existed deep networks. Deep acrivation featues are extracted from several pre-trained networks and employed in a traditional identification system. Moreover, a novel end-to-end convolutional deep architecture called "WBCsNet" is proposed and built from scratch. Finally, a limited balanced WBCs dataset classification is performed through the WBCsNet as a pre-trained network. During our experiments, three different public WBCs datasets (2551 images) have been used which contain 5 healthy WBCs types. The overall system accuracy achieved by the proposed WBCsNet is (96.1%) which is more than different transfer learning approaches or even the previous traditional identification system. We also present features visualization for the WBCsNet activation which reflects higher response than the pre-trained activated one. a novel WBCs identification system based on deep learning theory is proposed and a high performance WBCsNet can be employed as a pre-trained network. Copyright © 2017. Published by Elsevier B.V.

Updating algal evolutionary relationships through plastid genome sequencing: did alveolate plastids emerge through endosymbiosis of an ochrophyte?

PubMed

Ševčíková, Tereza; Horák, Aleš; Klimeš, Vladimír; Zbránková, Veronika; Demir-Hilton, Elif; Sudek, Sebastian; Jenkins, Jerry; Schmutz, Jeremy; Přibyl, Pavel; Fousek, Jan; Vlček, Čestmír; Lang, B Franz; Oborník, Miroslav; Worden, Alexandra Z; Eliáš, Marek

2015-05-28

Algae with secondary plastids of a red algal origin, such as ochrophytes (photosynthetic stramenopiles), are diverse and ecologically important, yet their evolutionary history remains controversial. We sequenced plastid genomes of two ochrophytes, Ochromonas sp. CCMP1393 (Chrysophyceae) and Trachydiscus minutus (Eustigmatophyceae). A shared split of the clpC gene as well as phylogenomic analyses of concatenated protein sequences demonstrated that chrysophytes and eustigmatophytes form a clade, the Limnista, exhibiting an unexpectedly elevated rate of plastid gene evolution. Our analyses also indicate that the root of the ochrophyte phylogeny falls between the recently redefined Khakista and Phaeista assemblages. Taking advantage of the expanded sampling of plastid genome sequences, we revisited the phylogenetic position of the plastid of Vitrella brassicaformis, a member of Alveolata with the least derived plastid genome known for the whole group. The results varied depending on the dataset and phylogenetic method employed, but suggested that the Vitrella plastids emerged from a deep ochrophyte lineage rather than being derived vertically from a hypothetical plastid-bearing common ancestor of alveolates and stramenopiles. Thus, we hypothesize that the plastid in Vitrella, and potentially in other alveolates, may have been acquired by an endosymbiosis of an early ochrophyte.

Speaker emotion recognition: from classical classifiers to deep neural networks

NASA Astrophysics Data System (ADS)

Mezghani, Eya; Charfeddine, Maha; Nicolas, Henri; Ben Amar, Chokri

2018-04-01

Speaker emotion recognition is considered among the most challenging tasks in recent years. In fact, automatic systems for security, medicine or education can be improved when considering the speech affective state. In this paper, a twofold approach for speech emotion classification is proposed. At the first side, a relevant set of features is adopted, and then at the second one, numerous supervised training techniques, involving classic methods as well as deep learning, are experimented. Experimental results indicate that deep architecture can improve classification performance on two affective databases, the Berlin Dataset of Emotional Speech and the SAVEE Dataset Surrey Audio-Visual Expressed Emotion.

Chromatin accessibility prediction via a hybrid deep convolutional neural network.

PubMed

Liu, Qiao; Xia, Fei; Yin, Qijin; Jiang, Rui

2018-03-01

A majority of known genetic variants associated with human-inherited diseases lie in non-coding regions that lack adequate interpretation, making it indispensable to systematically discover functional sites at the whole genome level and precisely decipher their implications in a comprehensive manner. Although computational approaches have been complementing high-throughput biological experiments towards the annotation of the human genome, it still remains a big challenge to accurately annotate regulatory elements in the context of a specific cell type via automatic learning of the DNA sequence code from large-scale sequencing data. Indeed, the development of an accurate and interpretable model to learn the DNA sequence signature and further enable the identification of causative genetic variants has become essential in both genomic and genetic studies. We proposed Deopen, a hybrid framework mainly based on a deep convolutional neural network, to automatically learn the regulatory code of DNA sequences and predict chromatin accessibility. In a series of comparison with existing methods, we show the superior performance of our model in not only the classification of accessible regions against background sequences sampled at random, but also the regression of DNase-seq signals. Besides, we further visualize the convolutional kernels and show the match of identified sequence signatures and known motifs. We finally demonstrate the sensitivity of our model in finding causative noncoding variants in the analysis of a breast cancer dataset. We expect to see wide applications of Deopen with either public or in-house chromatin accessibility data in the annotation of the human genome and the identification of non-coding variants associated with diseases. Deopen is freely available at https://github.com/kimmo1019/Deopen. ruijiang@tsinghua.edu.cn. Supplementary data are available at Bioinformatics online. © The Author (2017). Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Sequence2Vec: a novel embedding approach for modeling transcription factor binding affinity landscape.

PubMed

Dai, Hanjun; Umarov, Ramzan; Kuwahara, Hiroyuki; Li, Yu; Song, Le; Gao, Xin

2017-11-15

An accurate characterization of transcription factor (TF)-DNA affinity landscape is crucial to a quantitative understanding of the molecular mechanisms underpinning endogenous gene regulation. While recent advances in biotechnology have brought the opportunity for building binding affinity prediction methods, the accurate characterization of TF-DNA binding affinity landscape still remains a challenging problem. Here we propose a novel sequence embedding approach for modeling the transcription factor binding affinity landscape. Our method represents DNA binding sequences as a hidden Markov model which captures both position specific information and long-range dependency in the sequence. A cornerstone of our method is a novel message passing-like embedding algorithm, called Sequence2Vec, which maps these hidden Markov models into a common nonlinear feature space and uses these embedded features to build a predictive model. Our method is a novel combination of the strength of probabilistic graphical models, feature space embedding and deep learning. We conducted comprehensive experiments on over 90 large-scale TF-DNA datasets which were measured by different high-throughput experimental technologies. Sequence2Vec outperforms alternative machine learning methods as well as the state-of-the-art binding affinity prediction methods. Our program is freely available at https://github.com/ramzan1990/sequence2vec. xin.gao@kaust.edu.sa or lsong@cc.gatech.edu. Supplementary data are available at Bioinformatics online. © The Author(s) 2017. Published by Oxford University Press.

Viral to metazoan marine plankton nucleotide sequences from the Tara Oceans expedition

PubMed Central

Alberti, Adriana; Poulain, Julie; Engelen, Stefan; Labadie, Karine; Romac, Sarah; Ferrera, Isabel; Albini, Guillaume; Aury, Jean-Marc; Belser, Caroline; Bertrand, Alexis; Cruaud, Corinne; Da Silva, Corinne; Dossat, Carole; Gavory, Frédérick; Gas, Shahinaz; Guy, Julie; Haquelle, Maud; Jacoby, E'krame; Jaillon, Olivier; Lemainque, Arnaud; Pelletier, Eric; Samson, Gaëlle; Wessner, Mark; Bazire, Pascal; Beluche, Odette; Bertrand, Laurie; Besnard-Gonnet, Marielle; Bordelais, Isabelle; Boutard, Magali; Dubois, Maria; Dumont, Corinne; Ettedgui, Evelyne; Fernandez, Patricia; Garcia, Espérance; Aiach, Nathalie Giordanenco; Guerin, Thomas; Hamon, Chadia; Brun, Elodie; Lebled, Sandrine; Lenoble, Patricia; Louesse, Claudine; Mahieu, Eric; Mairey, Barbara; Martins, Nathalie; Megret, Catherine; Milani, Claire; Muanga, Jacqueline; Orvain, Céline; Payen, Emilie; Perroud, Peggy; Petit, Emmanuelle; Robert, Dominique; Ronsin, Murielle; Vacherie, Benoit; Acinas, Silvia G.; Royo-Llonch, Marta; Cornejo-Castillo, Francisco M.; Logares, Ramiro; Fernández-Gómez, Beatriz; Bowler, Chris; Cochrane, Guy; Amid, Clara; Hoopen, Petra Ten; De Vargas, Colomban; Grimsley, Nigel; Desgranges, Elodie; Kandels-Lewis, Stefanie; Ogata, Hiroyuki; Poulton, Nicole; Sieracki, Michael E.; Stepanauskas, Ramunas; Sullivan, Matthew B.; Brum, Jennifer R.; Duhaime, Melissa B.; Poulos, Bonnie T.; Hurwitz, Bonnie L.; Acinas, Silvia G.; Bork, Peer; Boss, Emmanuel; Bowler, Chris; De Vargas, Colomban; Follows, Michael; Gorsky, Gabriel; Grimsley, Nigel; Hingamp, Pascal; Iudicone, Daniele; Jaillon, Olivier; Kandels-Lewis, Stefanie; Karp-Boss, Lee; Karsenti, Eric; Not, Fabrice; Ogata, Hiroyuki; Pesant, Stéphane; Raes, Jeroen; Sardet, Christian; Sieracki, Michael E.; Speich, Sabrina; Stemmann, Lars; Sullivan, Matthew B.; Sunagawa, Shinichi; Wincker, Patrick; Pesant, Stéphane; Karsenti, Eric; Wincker, Patrick

2017-01-01

A unique collection of oceanic samples was gathered by the Tara Oceans expeditions (2009–2013), targeting plankton organisms ranging from viruses to metazoans, and providing rich environmental context measurements. Thanks to recent advances in the field of genomics, extensive sequencing has been performed for a deep genomic analysis of this huge collection of samples. A strategy based on different approaches, such as metabarcoding, metagenomics, single-cell genomics and metatranscriptomics, has been chosen for analysis of size-fractionated plankton communities. Here, we provide detailed procedures applied for genomic data generation, from nucleic acids extraction to sequence production, and we describe registries of genomics datasets available at the European Nucleotide Archive (ENA, www.ebi.ac.uk/ena). The association of these metadata to the experimental procedures applied for their generation will help the scientific community to access these data and facilitate their analysis. This paper complements other efforts to provide a full description of experiments and open science resources generated from the Tara Oceans project, further extending their value for the study of the world’s planktonic ecosystems. PMID:28763055

Viral to metazoan marine plankton nucleotide sequences from the Tara Oceans expedition.

PubMed

Alberti, Adriana; Poulain, Julie; Engelen, Stefan; Labadie, Karine; Romac, Sarah; Ferrera, Isabel; Albini, Guillaume; Aury, Jean-Marc; Belser, Caroline; Bertrand, Alexis; Cruaud, Corinne; Da Silva, Corinne; Dossat, Carole; Gavory, Frédérick; Gas, Shahinaz; Guy, Julie; Haquelle, Maud; Jacoby, E'krame; Jaillon, Olivier; Lemainque, Arnaud; Pelletier, Eric; Samson, Gaëlle; Wessner, Mark; Acinas, Silvia G; Royo-Llonch, Marta; Cornejo-Castillo, Francisco M; Logares, Ramiro; Fernández-Gómez, Beatriz; Bowler, Chris; Cochrane, Guy; Amid, Clara; Hoopen, Petra Ten; De Vargas, Colomban; Grimsley, Nigel; Desgranges, Elodie; Kandels-Lewis, Stefanie; Ogata, Hiroyuki; Poulton, Nicole; Sieracki, Michael E; Stepanauskas, Ramunas; Sullivan, Matthew B; Brum, Jennifer R; Duhaime, Melissa B; Poulos, Bonnie T; Hurwitz, Bonnie L; Pesant, Stéphane; Karsenti, Eric; Wincker, Patrick

2017-08-01

A unique collection of oceanic samples was gathered by the Tara Oceans expeditions (2009-2013), targeting plankton organisms ranging from viruses to metazoans, and providing rich environmental context measurements. Thanks to recent advances in the field of genomics, extensive sequencing has been performed for a deep genomic analysis of this huge collection of samples. A strategy based on different approaches, such as metabarcoding, metagenomics, single-cell genomics and metatranscriptomics, has been chosen for analysis of size-fractionated plankton communities. Here, we provide detailed procedures applied for genomic data generation, from nucleic acids extraction to sequence production, and we describe registries of genomics datasets available at the European Nucleotide Archive (ENA, www.ebi.ac.uk/ena). The association of these metadata to the experimental procedures applied for their generation will help the scientific community to access these data and facilitate their analysis. This paper complements other efforts to provide a full description of experiments and open science resources generated from the Tara Oceans project, further extending their value for the study of the world's planktonic ecosystems.

Rapid and accurate intraoperative pathological diagnosis by artificial intelligence with deep learning technology.

PubMed

Zhang, Jing; Song, Yanlin; Xia, Fan; Zhu, Chenjing; Zhang, Yingying; Song, Wenpeng; Xu, Jianguo; Ma, Xuelei

2017-09-01

Frozen section is widely used for intraoperative pathological diagnosis (IOPD), which is essential for intraoperative decision making. However, frozen section suffers from some drawbacks, such as time consuming and high misdiagnosis rate. Recently, artificial intelligence (AI) with deep learning technology has shown bright future in medicine. We hypothesize that AI with deep learning technology could help IOPD, with a computer trained by a dataset of intraoperative lesion images. Evidences supporting our hypothesis included the successful use of AI with deep learning technology in diagnosing skin cancer, and the developed method of deep-learning algorithm. Large size of the training dataset is critical to increase the diagnostic accuracy. The performance of the trained machine could be tested by new images before clinical use. Real-time diagnosis, easy to use and potential high accuracy were the advantages of AI for IOPD. In sum, AI with deep learning technology is a promising method to help rapid and accurate IOPD. Copyright © 2017 Elsevier Ltd. All rights reserved.

Achieving high confidence protein annotations in a sea of unknowns

NASA Astrophysics Data System (ADS)

Timmins-Schiffman, E.; May, D. H.; Noble, W. S.; Nunn, B. L.; Mikan, M.; Harvey, H. R.

2016-02-01

Increased sensitivity of mass spectrometry (MS) technology allows deep and broad insight into community functional analyses. Metaproteomics holds the promise to reveal functional responses of natural microbial communities, whereas metagenomics alone can only hint at potential functions. The complex datasets resulting from ocean MS have the potential to inform diverse realms of the biological, chemical, and physical ocean sciences, yet the extent of bacterial functional diversity and redundancy has not been fully explored. To take advantage of these impressive datasets, we need a clear bioinformatics pipeline for metaproteomics peptide identification and annotation with a database that can provide confident identifications. Researchers must consider whether it is sufficient to leverage the vast quantities of available ocean sequence data or if they must invest in site-specific metagenomic sequencing. We have sequenced, to our knowledge, the first western arctic metagenomes from the Bering Strait and the Chukchi Sea. We have addressed the long standing question: Is a metagenome required to accurately complete metaproteomics and assess the biological distribution of metabolic functions controlling nutrient acquisition in the ocean? Two different protein databases were constructed from 1) a site-specific metagenome and 2) subarctic/arctic groups available in NCBI's non-redundant database. Multiple proteomic search strategies were employed, against each individual database and against both databases combined, to determine the algorithm and approach that yielded the balance of high sensitivity and confident identification. Results yielded over 8200 confidently identified proteins. Our comparison of these results allows us to quantify the utility of investing resources in a metagenome versus using the constantly expanding and immediately available public databases for metaproteomic studies.

Leveraging long read sequencing from a single individual to provide a comprehensive resource for benchmarking variant calling methods

PubMed Central

Mu, John C.; Tootoonchi Afshar, Pegah; Mohiyuddin, Marghoob; Chen, Xi; Li, Jian; Bani Asadi, Narges; Gerstein, Mark B.; Wong, Wing H.; Lam, Hugo Y. K.

2015-01-01

A high-confidence, comprehensive human variant set is critical in assessing accuracy of sequencing algorithms, which are crucial in precision medicine based on high-throughput sequencing. Although recent works have attempted to provide such a resource, they still do not encompass all major types of variants including structural variants (SVs). Thus, we leveraged the massive high-quality Sanger sequences from the HuRef genome to construct by far the most comprehensive gold set of a single individual, which was cross validated with deep Illumina sequencing, population datasets, and well-established algorithms. It was a necessary effort to completely reanalyze the HuRef genome as its previously published variants were mostly reported five years ago, suffering from compatibility, organization, and accuracy issues that prevent their direct use in benchmarking. Our extensive analysis and validation resulted in a gold set with high specificity and sensitivity. In contrast to the current gold sets of the NA12878 or HS1011 genomes, our gold set is the first that includes small variants, deletion SVs and insertion SVs up to a hundred thousand base-pairs. We demonstrate the utility of our HuRef gold set to benchmark several published SV detection tools. PMID:26412485

Deep nirS amplicon sequencing of San Francisco Bay sediments enables prediction of geography and environmental conditions from denitrifying community composition.

PubMed

Lee, Jessica A; Francis, Christopher A

2017-12-01

Denitrification is a dominant nitrogen loss process in the sediments of San Francisco Bay. In this study, we sought to understand the ecology of denitrifying bacteria by using next-generation sequencing (NGS) to survey the diversity of a denitrification functional gene, nirS (encoding cytchrome-cd 1 nitrite reductase), along the salinity gradient of San Francisco Bay over the course of a year. We compared our dataset to a library of nirS sequences obtained previously from the same samples by standard PCR cloning and Sanger sequencing, and showed that both methods similarly demonstrated geography, salinity and, to a lesser extent, nitrogen, to be strong determinants of community composition. Furthermore, the depth afforded by NGS enabled novel techniques for measuring the association between environment and community composition. We used Random Forests modelling to demonstrate that the site and salinity of a sample could be predicted from its nirS sequences, and to identify indicator taxa associated with those environmental characteristics. This work contributes significantly to our understanding of the distribution and dynamics of denitrifying communities in San Francisco Bay, and provides valuable tools for the further study of this key N-cycling guild in all estuarine systems. © 2017 Society for Applied Microbiology and John Wiley & Sons Ltd.

Cough event classification by pretrained deep neural network.

PubMed

Liu, Jia-Ming; You, Mingyu; Wang, Zheng; Li, Guo-Zheng; Xu, Xianghuai; Qiu, Zhongmin

2015-01-01

Cough is an essential symptom in respiratory diseases. In the measurement of cough severity, an accurate and objective cough monitor is expected by respiratory disease society. This paper aims to introduce a better performed algorithm, pretrained deep neural network (DNN), to the cough classification problem, which is a key step in the cough monitor. The deep neural network models are built from two steps, pretrain and fine-tuning, followed by a Hidden Markov Model (HMM) decoder to capture tamporal information of the audio signals. By unsupervised pretraining a deep belief network, a good initialization for a deep neural network is learned. Then the fine-tuning step is a back propogation tuning the neural network so that it can predict the observation probability associated with each HMM states, where the HMM states are originally achieved by force-alignment with a Gaussian Mixture Model Hidden Markov Model (GMM-HMM) on the training samples. Three cough HMMs and one noncough HMM are employed to model coughs and noncoughs respectively. The final decision is made based on viterbi decoding algorihtm that generates the most likely HMM sequence for each sample. A sample is labeled as cough if a cough HMM is found in the sequence. The experiments were conducted on a dataset that was collected from 22 patients with respiratory diseases. Patient dependent (PD) and patient independent (PI) experimental settings were used to evaluate the models. Five criteria, sensitivity, specificity, F1, macro average and micro average are shown to depict different aspects of the models. From overall evaluation criteria, the DNN based methods are superior to traditional GMM-HMM based method on F1 and micro average with maximal 14% and 11% error reduction in PD and 7% and 10% in PI, meanwhile keep similar performances on macro average. They also surpass GMM-HMM model on specificity with maximal 14% error reduction on both PD and PI. In this paper, we tried pretrained deep neural network in cough classification problem. Our results showed that comparing with the conventional GMM-HMM framework, the HMM-DNN could get better overall performance on cough classification task.

Exploring fungal diversity in deep-sea sediments from Okinawa Trough using high-throughput Illumina sequencing

NASA Astrophysics Data System (ADS)

Zhang, Xiao-Yong; Wang, Guang-Hua; Xu, Xin-Ya; Nong, Xu-Hua; Wang, Jie; Amin, Muhammad; Qi, Shu-Hua

2016-10-01

The present study investigated the fungal diversity in four different deep-sea sediments from Okinawa Trough using high-throughput Illumina sequencing of the nuclear ribosomal internal transcribed spacer-1 (ITS1). A total of 40,297 fungal ITS1 sequences clustered into 420 operational taxonomic units (OTUs) with 97% sequence similarity and 170 taxa were recovered from these sediments. Most ITS1 sequences (78%) belonged to the phylum Ascomycota, followed by Basidiomycota (17.3%), Zygomycota (1.5%) and Chytridiomycota (0.8%), and a small proportion (2.4%) belonged to unassigned fungal phyla. Compared with previous studies on fungal diversity of sediments from deep-sea environments by culture-dependent approach and clone library analysis, the present result suggested that Illumina sequencing had been dramatically accelerating the discovery of fungal community of deep-sea sediments. Furthermore, our results revealed that Sordariomycetes was the most diverse and abundant fungal class in this study, challenging the traditional view that the diversity of Sordariomycetes phylotypes was low in the deep-sea environments. In addition, more than 12 taxa accounted for 21.5% sequences were found to be rarely reported as deep-sea fungi, suggesting the deep-sea sediments from Okinawa Trough harbored a plethora of different fungal communities compared with other deep-sea environments. To our knowledge, this study is the first exploration of the fungal diversity in deep-sea sediments from Okinawa Trough using high-throughput Illumina sequencing.

Deep learning in the small sample size setting: cascaded feed forward neural networks for medical image segmentation

NASA Astrophysics Data System (ADS)

Gaonkar, Bilwaj; Hovda, David; Martin, Neil; Macyszyn, Luke

2016-03-01

Deep Learning, refers to large set of neural network based algorithms, have emerged as promising machine- learning tools in the general imaging and computer vision domains. Convolutional neural networks (CNNs), a specific class of deep learning algorithms, have been extremely effective in object recognition and localization in natural images. A characteristic feature of CNNs, is the use of a locally connected multi layer topology that is inspired by the animal visual cortex (the most powerful vision system in existence). While CNNs, perform admirably in object identification and localization tasks, typically require training on extremely large datasets. Unfortunately, in medical image analysis, large datasets are either unavailable or are extremely expensive to obtain. Further, the primary tasks in medical imaging are organ identification and segmentation from 3D scans, which are different from the standard computer vision tasks of object recognition. Thus, in order to translate the advantages of deep learning to medical image analysis, there is a need to develop deep network topologies and training methodologies, that are geared towards medical imaging related tasks and can work in a setting where dataset sizes are relatively small. In this paper, we present a technique for stacked supervised training of deep feed forward neural networks for segmenting organs from medical scans. Each `neural network layer' in the stack is trained to identify a sub region of the original image, that contains the organ of interest. By layering several such stacks together a very deep neural network is constructed. Such a network can be used to identify extremely small regions of interest in extremely large images, inspite of a lack of clear contrast in the signal or easily identifiable shape characteristics. What is even more intriguing is that the network stack achieves accurate segmentation even when it is trained on a single image with manually labelled ground truth. We validate this approach,using a publicly available head and neck CT dataset. We also show that a deep neural network of similar depth, if trained directly using backpropagation, cannot acheive the tasks achieved using our layer wise training paradigm.

Identifying and mitigating batch effects in whole genome sequencing data.

PubMed

Tom, Jennifer A; Reeder, Jens; Forrest, William F; Graham, Robert R; Hunkapiller, Julie; Behrens, Timothy W; Bhangale, Tushar R

2017-07-24

Large sample sets of whole genome sequencing with deep coverage are being generated, however assembling datasets from different sources inevitably introduces batch effects. These batch effects are not well understood and can be due to changes in the sequencing protocol or bioinformatics tools used to process the data. No systematic algorithms or heuristics exist to detect and filter batch effects or remove associations impacted by batch effects in whole genome sequencing data. We describe key quality metrics, provide a freely available software package to compute them, and demonstrate that identification of batch effects is aided by principal components analysis of these metrics. To mitigate batch effects, we developed new site-specific filters that identified and removed variants that falsely associated with the phenotype due to batch effect. These include filtering based on: a haplotype based genotype correction, a differential genotype quality test, and removing sites with missing genotype rate greater than 30% after setting genotypes with quality scores less than 20 to missing. This method removed 96.1% of unconfirmed genome-wide significant SNP associations and 97.6% of unconfirmed genome-wide significant indel associations. We performed analyses to demonstrate that: 1) These filters impacted variants known to be disease associated as 2 out of 16 confirmed associations in an AMD candidate SNP analysis were filtered, representing a reduction in power of 12.5%, 2) In the absence of batch effects, these filters removed only a small proportion of variants across the genome (type I error rate of 3%), and 3) in an independent dataset, the method removed 90.2% of unconfirmed genome-wide SNP associations and 89.8% of unconfirmed genome-wide indel associations. Researchers currently do not have effective tools to identify and mitigate batch effects in whole genome sequencing data. We developed and validated methods and filters to address this deficiency.

RNA-protein binding motifs mining with a new hybrid deep learning based cross-domain knowledge integration approach.

PubMed

Pan, Xiaoyong; Shen, Hong-Bin

2017-02-28

RNAs play key roles in cells through the interactions with proteins known as the RNA-binding proteins (RBP) and their binding motifs enable crucial understanding of the post-transcriptional regulation of RNAs. How the RBPs correctly recognize the target RNAs and why they bind specific positions is still far from clear. Machine learning-based algorithms are widely acknowledged to be capable of speeding up this process. Although many automatic tools have been developed to predict the RNA-protein binding sites from the rapidly growing multi-resource data, e.g. sequence, structure, their domain specific features and formats have posed significant computational challenges. One of current difficulties is that the cross-source shared common knowledge is at a higher abstraction level beyond the observed data, resulting in a low efficiency of direct integration of observed data across domains. The other difficulty is how to interpret the prediction results. Existing approaches tend to terminate after outputting the potential discrete binding sites on the sequences, but how to assemble them into the meaningful binding motifs is a topic worth of further investigation. In viewing of these challenges, we propose a deep learning-based framework (iDeep) by using a novel hybrid convolutional neural network and deep belief network to predict the RBP interaction sites and motifs on RNAs. This new protocol is featured by transforming the original observed data into a high-level abstraction feature space using multiple layers of learning blocks, where the shared representations across different domains are integrated. To validate our iDeep method, we performed experiments on 31 large-scale CLIP-seq datasets, and our results show that by integrating multiple sources of data, the average AUC can be improved by 8% compared to the best single-source-based predictor; and through cross-domain knowledge integration at an abstraction level, it outperforms the state-of-the-art predictors by 6%. Besides the overall enhanced prediction performance, the convolutional neural network module embedded in iDeep is also able to automatically capture the interpretable binding motifs for RBPs. Large-scale experiments demonstrate that these mined binding motifs agree well with the experimentally verified results, suggesting iDeep is a promising approach in the real-world applications. The iDeep framework not only can achieve promising performance than the state-of-the-art predictors, but also easily capture interpretable binding motifs. iDeep is available at http://www.csbio.sjtu.edu.cn/bioinf/iDeep.

Deep sequencing of the Camellia chekiangoleosa transcriptome revealed candidate genes for anthocyanin biosynthesis.

PubMed

Wang, Zhong-Wei; Jiang, Cong; Wen, Qiang; Wang, Na; Tao, Yuan-Yuan; Xu, Li-An

2014-03-15

Camellia chekiangoleosa is an important species of genus Camellia. It provides high-quality edible oil and has great ornamental value. The flowers are big and red which bloom between February and March. Flower pigmentation is closely related to the accumulation of anthocyanin. Although anthocyanin biosynthesis has been studied extensively in herbaceous plants, little molecular information on the anthocyanin biosynthesis pathway of C. chekiangoleosa is yet known. In the present study, a cDNA library was constructed to obtain detailed and general data from the flowers of C. chekiangoleosa. To explore the transcriptome of C. chekiangoleosa and investigate genes involved in anthocyanin biosynthesis, a 454 GS FLX Titanium platform was used to generate an EST dataset. About 46,279 sequences were obtained, and 24,593 (53.1%) were annotated. Using Blast search against the AGRIS, 1740 unigenes were found homologous to 599 Arabidopsis transcription factor genes. Based on the transcriptome dataset, nine anthocyanin biosynthesis pathway genes (PAL, CHS1, CHS2, CHS3, CHI, F3H, DFR, ANS, and UFGT) were identified and cloned. The spatio-temporal expression patterns of these genes were also analyzed using quantitative real-time polymerase chain reaction. The study results not only enrich the gene resource but also provide valuable information for further studies concerning anthocyanin biosynthesis. Copyright © 2014 Elsevier B.V. All rights reserved.

EchinoDB, an application for comparative transcriptomics of deeply-sampled clades of echinoderms.

PubMed

Janies, Daniel A; Witter, Zach; Linchangco, Gregorio V; Foltz, David W; Miller, Allison K; Kerr, Alexander M; Jay, Jeremy; Reid, Robert W; Wray, Gregory A

2016-01-22

One of our goals for the echinoderm tree of life project (http://echinotol.org) is to identify orthologs suitable for phylogenetic analysis from next-generation transcriptome data. The current dataset is the largest assembled for echinoderm phylogeny and transcriptomics. We used RNA-Seq to profile adult tissues from 42 echinoderm specimens from 24 orders and 37 families. In order to achieve sampling members of clades that span key evolutionary divergence, many of our exemplars were collected from deep and polar seas. A small fraction of the transcriptome data we produced is being used for phylogenetic reconstruction. Thus to make a larger dataset available to researchers with a wide variety of interests, we made a web-based application, EchinoDB (http://echinodb.uncc.edu). EchinoDB is a repository of orthologous transcripts from echinoderms that is searchable via keywords and sequence similarity. From transcripts we identified 749,397 clusters of orthologous loci. We have developed the information technology to manage and search the loci their annotations with respect to the Sea Urchin (Strongylocentrotus purpuratus) genome. Several users have already taken advantage of these data for spin-off projects in developmental biology, gene family studies, and neuroscience. We hope others will search EchinoDB to discover datasets relevant to a variety of additional questions in comparative biology.

A Description of the Clinical Proteomic Tumor Analysis Consortium (CPTAC) Common Data Analysis Pipeline

PubMed Central

Rudnick, Paul A.; Markey, Sanford P.; Roth, Jeri; Mirokhin, Yuri; Yan, Xinjian; Tchekhovskoi, Dmitrii V.; Edwards, Nathan J.; Thangudu, Ratna R.; Ketchum, Karen A.; Kinsinger, Christopher R.; Mesri, Mehdi; Rodriguez, Henry; Stein, Stephen E.

2016-01-01

The Clinical Proteomic Tumor Analysis Consortium (CPTAC) has produced large proteomics datasets from the mass spectrometric interrogation of tumor samples previously analyzed by The Cancer Genome Atlas (TCGA) program. The availability of the genomic and proteomic data is enabling proteogenomic study for both reference (i.e., contained in major sequence databases) and non-reference markers of cancer. The CPTAC labs have focused on colon, breast, and ovarian tissues in the first round of analyses; spectra from these datasets were produced from 2D LC-MS/MS analyses and represent deep coverage. To reduce the variability introduced by disparate data analysis platforms (e.g., software packages, versions, parameters, sequence databases, etc.), the CPTAC Common Data Analysis Platform (CDAP) was created. The CDAP produces both peptide-spectrum-match (PSM) reports and gene-level reports. The pipeline processes raw mass spectrometry data according to the following: (1) Peak-picking and quantitative data extraction, (2) database searching, (3) gene-based protein parsimony, and (4) false discovery rate (FDR)-based filtering. The pipeline also produces localization scores for the phosphopeptide enrichment studies using the PhosphoRS program. Quantitative information for each of the datasets is specific to the sample processing, with PSM and protein reports containing the spectrum-level or gene-level (“rolled-up”) precursor peak areas and spectral counts for label-free or reporter ion log-ratios for 4plex iTRAQ™. The reports are available in simple tab-delimited formats and, for the PSM-reports, in mzIdentML. The goal of the CDAP is to provide standard, uniform reports for all of the CPTAC data, enabling comparisons between different samples and cancer types as well as across the major ‘omics fields. PMID:26860878

Simulating Next-Generation Sequencing Datasets from Empirical Mutation and Sequencing Models

PubMed Central

Stephens, Zachary D.; Hudson, Matthew E.; Mainzer, Liudmila S.; Taschuk, Morgan; Weber, Matthew R.; Iyer, Ravishankar K.

2016-01-01

An obstacle to validating and benchmarking methods for genome analysis is that there are few reference datasets available for which the “ground truth” about the mutational landscape of the sample genome is known and fully validated. Additionally, the free and public availability of real human genome datasets is incompatible with the preservation of donor privacy. In order to better analyze and understand genomic data, we need test datasets that model all variants, reflecting known biology as well as sequencing artifacts. Read simulators can fulfill this requirement, but are often criticized for limited resemblance to true data and overall inflexibility. We present NEAT (NExt-generation sequencing Analysis Toolkit), a set of tools that not only includes an easy-to-use read simulator, but also scripts to facilitate variant comparison and tool evaluation. NEAT has a wide variety of tunable parameters which can be set manually on the default model or parameterized using real datasets. The software is freely available at github.com/zstephens/neat-genreads. PMID:27893777

SamSelect: a sample sequence selection algorithm for quorum planted motif search on large DNA datasets.

PubMed

Yu, Qiang; Wei, Dingbang; Huo, Hongwei

2018-06-18

Given a set of t n-length DNA sequences, q satisfying 0 < q ≤ 1, and l and d satisfying 0 ≤ d < l < n, the quorum planted motif search (qPMS) finds l-length strings that occur in at least qt input sequences with up to d mismatches and is mainly used to locate transcription factor binding sites in DNA sequences. Existing qPMS algorithms have been able to efficiently process small standard datasets (e.g., t = 20 and n = 600), but they are too time consuming to process large DNA datasets, such as ChIP-seq datasets that contain thousands of sequences or more. We analyze the effects of t and q on the time performance of qPMS algorithms and find that a large t or a small q causes a longer computation time. Based on this information, we improve the time performance of existing qPMS algorithms by selecting a sample sequence set D' with a small t and a large q from the large input dataset D and then executing qPMS algorithms on D'. A sample sequence selection algorithm named SamSelect is proposed. The experimental results on both simulated and real data show (1) that SamSelect can select D' efficiently and (2) that the qPMS algorithms executed on D' can find implanted or real motifs in a significantly shorter time than when executed on D. We improve the ability of existing qPMS algorithms to process large DNA datasets from the perspective of selecting high-quality sample sequence sets so that the qPMS algorithms can find motifs in a short time in the selected sample sequence set D', rather than take an unfeasibly long time to search the original sequence set D. Our motif discovery method is an approximate algorithm.

Quantitative phenotyping via deep barcode sequencing.

PubMed

Smith, Andrew M; Heisler, Lawrence E; Mellor, Joseph; Kaper, Fiona; Thompson, Michael J; Chee, Mark; Roth, Frederick P; Giaever, Guri; Nislow, Corey

2009-10-01

Next-generation DNA sequencing technologies have revolutionized diverse genomics applications, including de novo genome sequencing, SNP detection, chromatin immunoprecipitation, and transcriptome analysis. Here we apply deep sequencing to genome-scale fitness profiling to evaluate yeast strain collections in parallel. This method, Barcode analysis by Sequencing, or "Bar-seq," outperforms the current benchmark barcode microarray assay in terms of both dynamic range and throughput. When applied to a complex chemogenomic assay, Bar-seq quantitatively identifies drug targets, with performance superior to the benchmark microarray assay. We also show that Bar-seq is well-suited for a multiplex format. We completely re-sequenced and re-annotated the yeast deletion collection using deep sequencing, found that approximately 20% of the barcodes and common priming sequences varied from expectation, and used this revised list of barcode sequences to improve data quality. Together, this new assay and analysis routine provide a deep-sequencing-based toolkit for identifying gene-environment interactions on a genome-wide scale.

Deep Sequencing to Identify the Causes of Viral Encephalitis

PubMed Central

Chan, Benjamin K.; Wilson, Theodore; Fischer, Kael F.; Kriesel, John D.

2014-01-01

Deep sequencing allows for a rapid, accurate characterization of microbial DNA and RNA sequences in many types of samples. Deep sequencing (also called next generation sequencing or NGS) is being developed to assist with the diagnosis of a wide variety of infectious diseases. In this study, seven frozen brain samples from deceased subjects with recent encephalitis were investigated. RNA from each sample was extracted, randomly reverse transcribed and sequenced. The sequence analysis was performed in a blinded fashion and confirmed with pathogen-specific PCR. This analysis successfully identified measles virus sequences in two brain samples and herpes simplex virus type-1 sequences in three brain samples. No pathogen was identified in the other two brain specimens. These results were concordant with pathogen-specific PCR and partially concordant with prior neuropathological examinations, demonstrating that deep sequencing can accurately identify viral infections in frozen brain tissue. PMID:24699691

Diagnosis and prediction of periodontally compromised teeth using a deep learning-based convolutional neural network algorithm.

PubMed

Lee, Jae-Hong; Kim, Do-Hyung; Jeong, Seong-Nyum; Choi, Seong-Ho

2018-04-01

The aim of the current study was to develop a computer-assisted detection system based on a deep convolutional neural network (CNN) algorithm and to evaluate the potential usefulness and accuracy of this system for the diagnosis and prediction of periodontally compromised teeth (PCT). Combining pretrained deep CNN architecture and a self-trained network, periapical radiographic images were used to determine the optimal CNN algorithm and weights. The diagnostic and predictive accuracy, sensitivity, specificity, positive predictive value, negative predictive value, receiver operating characteristic (ROC) curve, area under the ROC curve, confusion matrix, and 95% confidence intervals (CIs) were calculated using our deep CNN algorithm, based on a Keras framework in Python. The periapical radiographic dataset was split into training (n=1,044), validation (n=348), and test (n=348) datasets. With the deep learning algorithm, the diagnostic accuracy for PCT was 81.0% for premolars and 76.7% for molars. Using 64 premolars and 64 molars that were clinically diagnosed as severe PCT, the accuracy of predicting extraction was 82.8% (95% CI, 70.1%-91.2%) for premolars and 73.4% (95% CI, 59.9%-84.0%) for molars. We demonstrated that the deep CNN algorithm was useful for assessing the diagnosis and predictability of PCT. Therefore, with further optimization of the PCT dataset and improvements in the algorithm, a computer-aided detection system can be expected to become an effective and efficient method of diagnosing and predicting PCT.

Deep Filter Banks for Texture Recognition, Description, and Segmentation.

PubMed

Cimpoi, Mircea; Maji, Subhransu; Kokkinos, Iasonas; Vedaldi, Andrea

Visual textures have played a key role in image understanding because they convey important semantics of images, and because texture representations that pool local image descriptors in an orderless manner have had a tremendous impact in diverse applications. In this paper we make several contributions to texture understanding. First, instead of focusing on texture instance and material category recognition, we propose a human-interpretable vocabulary of texture attributes to describe common texture patterns, complemented by a new describable texture dataset for benchmarking. Second, we look at the problem of recognizing materials and texture attributes in realistic imaging conditions, including when textures appear in clutter, developing corresponding benchmarks on top of the recently proposed OpenSurfaces dataset. Third, we revisit classic texture represenations, including bag-of-visual-words and the Fisher vectors, in the context of deep learning and show that these have excellent efficiency and generalization properties if the convolutional layers of a deep model are used as filter banks. We obtain in this manner state-of-the-art performance in numerous datasets well beyond textures, an efficient method to apply deep features to image regions, as well as benefit in transferring features from one domain to another.

Transposable elements in TDP-43-mediated neurodegenerative disorders.

PubMed

Li, Wanhe; Jin, Ying; Prazak, Lisa; Hammell, Molly; Dubnau, Josh

2012-01-01

Elevated expression of specific transposable elements (TEs) has been observed in several neurodegenerative disorders. TEs also can be active during normal neurogenesis. By mining a series of deep sequencing datasets of protein-RNA interactions and of gene expression profiles, we uncovered extensive binding of TE transcripts to TDP-43, an RNA-binding protein central to amyotrophic lateral sclerosis (ALS) and frontotemporal lobar degeneration (FTLD). Second, we find that association between TDP-43 and many of its TE targets is reduced in FTLD patients. Third, we discovered that a large fraction of the TEs to which TDP-43 binds become de-repressed in mouse TDP-43 disease models. We propose the hypothesis that TE mis-regulation contributes to TDP-43 related neurodegenerative diseases.

Scalable metagenomic taxonomy classification using a reference genome database

PubMed Central

Ames, Sasha K.; Hysom, David A.; Gardner, Shea N.; Lloyd, G. Scott; Gokhale, Maya B.; Allen, Jonathan E.

2013-01-01

Motivation: Deep metagenomic sequencing of biological samples has the potential to recover otherwise difficult-to-detect microorganisms and accurately characterize biological samples with limited prior knowledge of sample contents. Existing metagenomic taxonomic classification algorithms, however, do not scale well to analyze large metagenomic datasets, and balancing classification accuracy with computational efficiency presents a fundamental challenge. Results: A method is presented to shift computational costs to an off-line computation by creating a taxonomy/genome index that supports scalable metagenomic classification. Scalable performance is demonstrated on real and simulated data to show accurate classification in the presence of novel organisms on samples that include viruses, prokaryotes, fungi and protists. Taxonomic classification of the previously published 150 giga-base Tyrolean Iceman dataset was found to take <20 h on a single node 40 core large memory machine and provide new insights on the metagenomic contents of the sample. Availability: Software was implemented in C++ and is freely available at http://sourceforge.net/projects/lmat Contact: allen99@llnl.gov Supplementary information: Supplementary data are available at Bioinformatics online. PMID:23828782

EnzyNet: enzyme classification using 3D convolutional neural networks on spatial representation

PubMed Central

Amidi, Afshine; Megalooikonomou, Vasileios; Paragios, Nikos

2018-01-01

During the past decade, with the significant progress of computational power as well as ever-rising data availability, deep learning techniques became increasingly popular due to their excellent performance on computer vision problems. The size of the Protein Data Bank (PDB) has increased more than 15-fold since 1999, which enabled the expansion of models that aim at predicting enzymatic function via their amino acid composition. Amino acid sequence, however, is less conserved in nature than protein structure and therefore considered a less reliable predictor of protein function. This paper presents EnzyNet, a novel 3D convolutional neural networks classifier that predicts the Enzyme Commission number of enzymes based only on their voxel-based spatial structure. The spatial distribution of biochemical properties was also examined as complementary information. The two-layer architecture was investigated on a large dataset of 63,558 enzymes from the PDB and achieved an accuracy of 78.4% by exploiting only the binary representation of the protein shape. Code and datasets are available at https://github.com/shervinea/enzynet. PMID:29740518

EnzyNet: enzyme classification using 3D convolutional neural networks on spatial representation.

PubMed

Amidi, Afshine; Amidi, Shervine; Vlachakis, Dimitrios; Megalooikonomou, Vasileios; Paragios, Nikos; Zacharaki, Evangelia I

2018-01-01

During the past decade, with the significant progress of computational power as well as ever-rising data availability, deep learning techniques became increasingly popular due to their excellent performance on computer vision problems. The size of the Protein Data Bank (PDB) has increased more than 15-fold since 1999, which enabled the expansion of models that aim at predicting enzymatic function via their amino acid composition. Amino acid sequence, however, is less conserved in nature than protein structure and therefore considered a less reliable predictor of protein function. This paper presents EnzyNet, a novel 3D convolutional neural networks classifier that predicts the Enzyme Commission number of enzymes based only on their voxel-based spatial structure. The spatial distribution of biochemical properties was also examined as complementary information. The two-layer architecture was investigated on a large dataset of 63,558 enzymes from the PDB and achieved an accuracy of 78.4% by exploiting only the binary representation of the protein shape. Code and datasets are available at https://github.com/shervinea/enzynet.

Completing the Legacy of Hubble's Wide/Deep Fields: An Aligned Complete Dataset of 1220 Orbits on the GOODS-N/CANDELS-N Region

NASA Astrophysics Data System (ADS)

Illingworth, Garth

2017-08-01

The GOODS-N/CANDELS-N region is second only to the GOODS-S/ECDF-S region in the extent of its HST and Spitzer coverage, making it a remarkable science resource. Yet of 1220 orbits of ACS and WFC3/IR imaging from 27 programs on the GOODS-N region, fully 42% of the total, about 520 orbits of imaging data from 22 programs, remains unavailable in MAST as a high-level science data product (HLSP). The GOODS-N region dataset is a key Legacy field ( 3 Msec from HST, 6 Msec from Spitzer, and 2 Msec from Chandra). We propose to deliver, with catalogs, HST ACS and WFC3/IR HLSPs to MAST for all 1220 orbits of GOODS-N data. We will also deliver HLSPs for the EGS, UDS and the COSMOS CANDELS regions, including new data not included to date. These four HLSPs, 2300 orbits of HST data ( 75% of a HST Cycle ), will add substantially to (1) our understanding of the build-up of galaxies to z 6 in the first Gyr during reionization, (2) the development of galaxies over the subsequent Gyr to the peak of the star formation rate in the universe at z 2-3, and (3) the transition at z<2 of early star-forming galaxies to the full splendor of the Hubble sequence. We can do this major AR Legacy program, having submitted a HLSP of ALL 2442 orbits of HST data on the GOODS- S region (>950 orbits new). The total volume of data in the GOODS-S Hubble Legacy Field (HLF-GOODS-S) is 5.8 Msec in 7211 exposures ( 70% of a HST cycle). The HLF-GOODS-S includes 4 new deep areas akin to the HUDF/XDF. The four proposed NEW Hubble Legacy Field datasets will complement the Frontier Field datasets and our recent HLF-GOODS-S and HUDF/XDF HLSP submissions. They will be cornerstones of Hubble's Legacy as the JWST era dawns.

PhyloPythiaS+: a self-training method for the rapid reconstruction of low-ranking taxonomic bins from metagenomes.

PubMed

Gregor, Ivan; Dröge, Johannes; Schirmer, Melanie; Quince, Christopher; McHardy, Alice C

2016-01-01

Background. Metagenomics is an approach for characterizing environmental microbial communities in situ, it allows their functional and taxonomic characterization and to recover sequences from uncultured taxa. This is often achieved by a combination of sequence assembly and binning, where sequences are grouped into 'bins' representing taxa of the underlying microbial community. Assignment to low-ranking taxonomic bins is an important challenge for binning methods as is scalability to Gb-sized datasets generated with deep sequencing techniques. One of the best available methods for species bins recovery from deep-branching phyla is the expert-trained PhyloPythiaS package, where a human expert decides on the taxa to incorporate in the model and identifies 'training' sequences based on marker genes directly from the sample. Due to the manual effort involved, this approach does not scale to multiple metagenome samples and requires substantial expertise, which researchers who are new to the area do not have. Results. We have developed PhyloPythiaS+, a successor to our PhyloPythia(S) software. The new (+) component performs the work previously done by the human expert. PhyloPythiaS+ also includes a new k-mer counting algorithm, which accelerated the simultaneous counting of 4-6-mers used for taxonomic binning 100-fold and reduced the overall execution time of the software by a factor of three. Our software allows to analyze Gb-sized metagenomes with inexpensive hardware, and to recover species or genera-level bins with low error rates in a fully automated fashion. PhyloPythiaS+ was compared to MEGAN, taxator-tk, Kraken and the generic PhyloPythiaS model. The results showed that PhyloPythiaS+ performs especially well for samples originating from novel environments in comparison to the other methods. Availability. PhyloPythiaS+ in a virtual machine is available for installation under Windows, Unix systems or OS X on: https://github.com/algbioi/ppsp/wiki.

ChemNet: A Transferable and Generalizable Deep Neural Network for Small-Molecule Property Prediction

DOE Office of Scientific and Technical Information (OSTI.GOV)

Goh, Garrett B.; Siegel, Charles M.; Vishnu, Abhinav

With access to large datasets, deep neural networks through representation learning have been able to identify patterns from raw data, achieving human-level accuracy in image and speech recognition tasks. However, in chemistry, availability of large standardized and labelled datasets is scarce, and with a multitude of chemical properties of interest, chemical data is inherently small and fragmented. In this work, we explore transfer learning techniques in conjunction with the existing Chemception CNN model, to create a transferable and generalizable deep neural network for small-molecule property prediction. Our latest model, ChemNet learns in a semi-supervised manner from inexpensive labels computed frommore » the ChEMBL database. When fine-tuned to the Tox21, HIV and FreeSolv dataset, which are 3 separate chemical tasks that ChemNet was not originally trained on, we demonstrate that ChemNet exceeds the performance of existing Chemception models, contemporary MLP models that trains on molecular fingerprints, and it matches the performance of the ConvGraph algorithm, the current state-of-the-art. Furthermore, as ChemNet has been pre-trained on a large diverse chemical database, it can be used as a universal “plug-and-play” deep neural network, which accelerates the deployment of deep neural networks for the prediction of novel small-molecule chemical properties.« less

iSS-PC: Identifying Splicing Sites via Physical-Chemical Properties Using Deep Sparse Auto-Encoder.

PubMed

Xu, Zhao-Chun; Wang, Peng; Qiu, Wang-Ren; Xiao, Xuan

2017-08-15

Gene splicing is one of the most significant biological processes in eukaryotic gene expression, such as RNA splicing, which can cause a pre-mRNA to produce one or more mature messenger RNAs containing the coded information with multiple biological functions. Thus, identifying splicing sites in DNA/RNA sequences is significant for both the bio-medical research and the discovery of new drugs. However, it is expensive and time consuming based only on experimental technique, so new computational methods are needed. To identify the splice donor sites and splice acceptor sites accurately and quickly, a deep sparse auto-encoder model with two hidden layers, called iSS-PC, was constructed based on minimum error law, in which we incorporated twelve physical-chemical properties of the dinucleotides within DNA into PseDNC to formulate given sequence samples via a battery of cross-covariance and auto-covariance transformations. In this paper, five-fold cross-validation test results based on the same benchmark data-sets indicated that the new predictor remarkably outperformed the existing prediction methods in this field. Furthermore, it is expected that many other related problems can be also studied by this approach. To implement classification accurately and quickly, an easy-to-use web-server for identifying slicing sites has been established for free access at: http://www.jci-bioinfo.cn/iSS-PC.

Lifelong learning of human actions with deep neural network self-organization.

PubMed

Parisi, German I; Tani, Jun; Weber, Cornelius; Wermter, Stefan

2017-12-01

Lifelong learning is fundamental in autonomous robotics for the acquisition and fine-tuning of knowledge through experience. However, conventional deep neural models for action recognition from videos do not account for lifelong learning but rather learn a batch of training data with a predefined number of action classes and samples. Thus, there is the need to develop learning systems with the ability to incrementally process available perceptual cues and to adapt their responses over time. We propose a self-organizing neural architecture for incrementally learning to classify human actions from video sequences. The architecture comprises growing self-organizing networks equipped with recurrent neurons for processing time-varying patterns. We use a set of hierarchically arranged recurrent networks for the unsupervised learning of action representations with increasingly large spatiotemporal receptive fields. Lifelong learning is achieved in terms of prediction-driven neural dynamics in which the growth and the adaptation of the recurrent networks are driven by their capability to reconstruct temporally ordered input sequences. Experimental results on a classification task using two action benchmark datasets show that our model is competitive with state-of-the-art methods for batch learning also when a significant number of sample labels are missing or corrupted during training sessions. Additional experiments show the ability of our model to adapt to non-stationary input avoiding catastrophic interference. Copyright © 2017 The Author(s). Published by Elsevier Ltd.. All rights reserved.

Metaxa: a software tool for automated detection and discrimination among ribosomal small subunit (12S/16S/18S) sequences of archaea, bacteria, eukaryotes, mitochondria, and chloroplasts in metagenomes and environmental sequencing datasets.

PubMed

Bengtsson, Johan; Eriksson, K Martin; Hartmann, Martin; Wang, Zheng; Shenoy, Belle Damodara; Grelet, Gwen-Aëlle; Abarenkov, Kessy; Petri, Anna; Rosenblad, Magnus Alm; Nilsson, R Henrik

2011-10-01

The ribosomal small subunit (SSU) rRNA gene has emerged as an important genetic marker for taxonomic identification in environmental sequencing datasets. In addition to being present in the nucleus of eukaryotes and the core genome of prokaryotes, the gene is also found in the mitochondria of eukaryotes and in the chloroplasts of photosynthetic eukaryotes. These three sets of genes are conceptually paralogous and should in most situations not be aligned and analyzed jointly. To identify the origin of SSU sequences in complex sequence datasets has hitherto been a time-consuming and largely manual undertaking. However, the present study introduces Metaxa ( http://microbiology.se/software/metaxa/ ), an automated software tool to extract full-length and partial SSU sequences from larger sequence datasets and assign them to an archaeal, bacterial, nuclear eukaryote, mitochondrial, or chloroplast origin. Using data from reference databases and from full-length organelle and organism genomes, we show that Metaxa detects and scores SSU sequences for origin with very low proportions of false positives and negatives. We believe that this tool will be useful in microbial and evolutionary ecology as well as in metagenomics.

Quantitative phenotyping via deep barcode sequencing

PubMed Central

Smith, Andrew M.; Heisler, Lawrence E.; Mellor, Joseph; Kaper, Fiona; Thompson, Michael J.; Chee, Mark; Roth, Frederick P.; Giaever, Guri; Nislow, Corey

2009-01-01

Next-generation DNA sequencing technologies have revolutionized diverse genomics applications, including de novo genome sequencing, SNP detection, chromatin immunoprecipitation, and transcriptome analysis. Here we apply deep sequencing to genome-scale fitness profiling to evaluate yeast strain collections in parallel. This method, Barcode analysis by Sequencing, or “Bar-seq,” outperforms the current benchmark barcode microarray assay in terms of both dynamic range and throughput. When applied to a complex chemogenomic assay, Bar-seq quantitatively identifies drug targets, with performance superior to the benchmark microarray assay. We also show that Bar-seq is well-suited for a multiplex format. We completely re-sequenced and re-annotated the yeast deletion collection using deep sequencing, found that ∼20% of the barcodes and common priming sequences varied from expectation, and used this revised list of barcode sequences to improve data quality. Together, this new assay and analysis routine provide a deep-sequencing-based toolkit for identifying gene–environment interactions on a genome-wide scale. PMID:19622793

Exploring Genetic Divergence in a Species-Rich Insect Genus Using 2790 DNA Barcodes

PubMed Central

Lin, Xiaolong; Stur, Elisabeth; Ekrem, Torbjørn

2015-01-01

DNA barcoding using a fragment of the mitochondrial cytochrome c oxidase subunit 1 gene (COI) has proven to be successful for species-level identification in many animal groups. However, most studies have been focused on relatively small datasets or on large datasets of taxonomically high-ranked groups. We explore the quality of DNA barcodes to delimit species in the diverse chironomid genus Tanytarsus (Diptera: Chironomidae) by using different analytical tools. The genus Tanytarsus is the most species-rich taxon of tribe Tanytarsini (Diptera: Chironomidae) with more than 400 species worldwide, some of which can be notoriously difficult to identify to species-level using morphology. Our dataset, based on sequences generated from own material and publicly available data in BOLD, consist of 2790 DNA barcodes with a fragment length of at least 500 base pairs. A neighbor joining tree of this dataset comprises 131 well separated clusters representing 121 morphological species of Tanytarsus: 77 named, 16 unnamed and 28 unidentified theoretical species. For our geographically widespread dataset, DNA barcodes unambiguously discriminate 94.6% of the Tanytarsus species recognized through prior morphological study. Deep intraspecific divergences exist in some species complexes, and need further taxonomic studies using appropriate nuclear markers as well as morphological and ecological data to be resolved. The DNA barcodes cluster into 120–242 molecular operational taxonomic units (OTUs) depending on whether Objective Clustering, Automatic Barcode Gap Discovery (ABGD), Generalized Mixed Yule Coalescent model (GMYC), Poisson Tree Process (PTP), subjective evaluation of the neighbor joining tree or Barcode Index Numbers (BINs) are used. We suggest that a 4–5% threshold is appropriate to delineate species of Tanytarsus non-biting midges. PMID:26406595

Reference datasets for 2-treatment, 2-sequence, 2-period bioequivalence studies.

PubMed

Schütz, Helmut; Labes, Detlew; Fuglsang, Anders

2014-11-01

It is difficult to validate statistical software used to assess bioequivalence since very few datasets with known results are in the public domain, and the few that are published are of moderate size and balanced. The purpose of this paper is therefore to introduce reference datasets of varying complexity in terms of dataset size and characteristics (balance, range, outlier presence, residual error distribution) for 2-treatment, 2-period, 2-sequence bioequivalence studies and to report their point estimates and 90% confidence intervals which companies can use to validate their installations. The results for these datasets were calculated using the commercial packages EquivTest, Kinetica, SAS and WinNonlin, and the non-commercial package R. The results of three of these packages mostly agree, but imbalance between sequences seems to provoke questionable results with one package, which illustrates well the need for proper software validation.

Contrasting morphological and DNA barcode-suggested species boundaries among shallow-water amphipod fauna from the southern European Atlantic coast.

PubMed

Lobo, Jorge; Ferreira, Maria S; Antunes, Ilisa C; Teixeira, Marcos A L; Borges, Luisa M S; Sousa, Ronaldo; Gomes, Pedro A; Costa, Maria Helena; Cunha, Marina R; Costa, Filipe O

2017-02-01

In this study we compared DNA barcode-suggested species boundaries with morphology-based species identifications in the amphipod fauna of the southern European Atlantic coast. DNA sequences of the cytochrome c oxidase subunit I barcode region (COI-5P) were generated for 43 morphospecies (178 specimens) collected along the Portuguese coast which, together with publicly available COI-5P sequences, produced a final dataset comprising 68 morphospecies and 295 sequences. Seventy-five BINs (Barcode Index Numbers) were assigned to these morphospecies, of which 48 were concordant (i.e., 1 BIN = 1 species), 8 were taxonomically discordant, and 19 were singletons. Twelve species had matching sequences (<2% distance) with conspecifics from distant locations (e.g., North Sea). Seven morphospecies were assigned to multiple, and highly divergent, BINs, including specimens of Corophium multisetosum (18% divergence) and Dexamine spiniventris (16% divergence), which originated from sampling locations on the west coast of Portugal (only about 36 and 250 km apart, respectively). We also found deep divergence (4%-22%) among specimens of seven species from Portugal compared to those from the North Sea and Italy. The detection of evolutionarily meaningful divergence among populations of several amphipod species from southern Europe reinforces the need for a comprehensive re-assessment of the diversity of this faunal group.

A combined deep-learning and deformable-model approach to fully automatic segmentation of the left ventricle in cardiac MRI.

PubMed

Avendi, M R; Kheradvar, Arash; Jafarkhani, Hamid

2016-05-01

Segmentation of the left ventricle (LV) from cardiac magnetic resonance imaging (MRI) datasets is an essential step for calculation of clinical indices such as ventricular volume and ejection fraction. In this work, we employ deep learning algorithms combined with deformable models to develop and evaluate a fully automatic LV segmentation tool from short-axis cardiac MRI datasets. The method employs deep learning algorithms to learn the segmentation task from the ground true data. Convolutional networks are employed to automatically detect the LV chamber in MRI dataset. Stacked autoencoders are used to infer the LV shape. The inferred shape is incorporated into deformable models to improve the accuracy and robustness of the segmentation. We validated our method using 45 cardiac MR datasets from the MICCAI 2009 LV segmentation challenge and showed that it outperforms the state-of-the art methods. Excellent agreement with the ground truth was achieved. Validation metrics, percentage of good contours, Dice metric, average perpendicular distance and conformity, were computed as 96.69%, 0.94, 1.81 mm and 0.86, versus those of 79.2-95.62%, 0.87-0.9, 1.76-2.97 mm and 0.67-0.78, obtained by other methods, respectively. Copyright © 2016 Elsevier B.V. All rights reserved.

Deep Learning for Lowtextured Image Matching

NASA Astrophysics Data System (ADS)

Kniaz, V. V.; Fedorenko, V. V.; Fomin, N. A.

2018-05-01

Low-textured objects pose challenges for an automatic 3D model reconstruction. Such objects are common in archeological applications of photogrammetry. Most of the common feature point descriptors fail to match local patches in featureless regions of an object. Hence, automatic documentation of the archeological process using Structure from Motion (SfM) methods is challenging. Nevertheless, such documentation is possible with the aid of a human operator. Deep learning-based descriptors have outperformed most of common feature point descriptors recently. This paper is focused on the development of a new Wide Image Zone Adaptive Robust feature Descriptor (WIZARD) based on the deep learning. We use a convolutional auto-encoder to compress discriminative features of a local path into a descriptor code. We build a codebook to perform point matching on multiple images. The matching is performed using the nearest neighbor search and a modified voting algorithm. We present a new "Multi-view Amphora" (Amphora) dataset for evaluation of point matching algorithms. The dataset includes images of an Ancient Greek vase found at Taman Peninsula in Southern Russia. The dataset provides color images, a ground truth 3D model, and a ground truth optical flow. We evaluated the WIZARD descriptor on the "Amphora" dataset to show that it outperforms the SIFT and SURF descriptors on the complex patch pairs.

BRAD, the genetics and genomics database for Brassica plants.

PubMed

Cheng, Feng; Liu, Shengyi; Wu, Jian; Fang, Lu; Sun, Silong; Liu, Bo; Li, Pingxia; Hua, Wei; Wang, Xiaowu

2011-10-13

Brassica species include both vegetable and oilseed crops, which are very important to the daily life of common human beings. Meanwhile, the Brassica species represent an excellent system for studying numerous aspects of plant biology, specifically for the analysis of genome evolution following polyploidy, so it is also very important for scientific research. Now, the genome of Brassica rapa has already been assembled, it is the time to do deep mining of the genome data. BRAD, the Brassica database, is a web-based resource focusing on genome scale genetic and genomic data for important Brassica crops. BRAD was built based on the first whole genome sequence and on further data analysis of the Brassica A genome species, Brassica rapa (Chiifu-401-42). It provides datasets, such as the complete genome sequence of B. rapa, which was de novo assembled from Illumina GA II short reads and from BAC clone sequences, predicted genes and associated annotations, non coding RNAs, transposable elements (TE), B. rapa genes' orthologous to those in A. thaliana, as well as genetic markers and linkage maps. BRAD offers useful searching and data mining tools, including search across annotation datasets, search for syntenic or non-syntenic orthologs, and to search the flanking regions of a certain target, as well as the tools of BLAST and Gbrowse. BRAD allows users to enter almost any kind of information, such as a B. rapa or A. thaliana gene ID, physical position or genetic marker. BRAD, a new database which focuses on the genetics and genomics of the Brassica plants has been developed, it aims at helping scientists and breeders to fully and efficiently use the information of genome data of Brassica plants. BRAD will be continuously updated and can be accessed through http://brassicadb.org.

DeepPap: Deep Convolutional Networks for Cervical Cell Classification.

PubMed

Zhang, Ling; Le Lu; Nogues, Isabella; Summers, Ronald M; Liu, Shaoxiong; Yao, Jianhua

2017-11-01

Automation-assisted cervical screening via Pap smear or liquid-based cytology (LBC) is a highly effective cell imaging based cancer detection tool, where cells are partitioned into "abnormal" and "normal" categories. However, the success of most traditional classification methods relies on the presence of accurate cell segmentations. Despite sixty years of research in this field, accurate segmentation remains a challenge in the presence of cell clusters and pathologies. Moreover, previous classification methods are only built upon the extraction of hand-crafted features, such as morphology and texture. This paper addresses these limitations by proposing a method to directly classify cervical cells-without prior segmentation-based on deep features, using convolutional neural networks (ConvNets). First, the ConvNet is pretrained on a natural image dataset. It is subsequently fine-tuned on a cervical cell dataset consisting of adaptively resampled image patches coarsely centered on the nuclei. In the testing phase, aggregation is used to average the prediction scores of a similar set of image patches. The proposed method is evaluated on both Pap smear and LBC datasets. Results show that our method outperforms previous algorithms in classification accuracy (98.3%), area under the curve (0.99) values, and especially specificity (98.3%), when applied to the Herlev benchmark Pap smear dataset and evaluated using five-fold cross validation. Similar superior performances are also achieved on the HEMLBC (H&E stained manual LBC) dataset. Our method is promising for the development of automation-assisted reading systems in primary cervical screening.

Deep Convolutional Neural Networks for Endotracheal Tube Position and X-ray Image Classification: Challenges and Opportunities.

PubMed

Lakhani, Paras

2017-08-01

The goal of this study is to evaluate the efficacy of deep convolutional neural networks (DCNNs) in differentiating subtle, intermediate, and more obvious image differences in radiography. Three different datasets were created, which included presence/absence of the endotracheal (ET) tube (n = 300), low/normal position of the ET tube (n = 300), and chest/abdominal radiographs (n = 120). The datasets were split into training, validation, and test. Both untrained and pre-trained deep neural networks were employed, including AlexNet and GoogLeNet classifiers, using the Caffe framework. Data augmentation was performed for the presence/absence and low/normal ET tube datasets. Receiver operating characteristic (ROC), area under the curves (AUC), and 95% confidence intervals were calculated. Statistical differences of the AUCs were determined using a non-parametric approach. The pre-trained AlexNet and GoogLeNet classifiers had perfect accuracy (AUC 1.00) in differentiating chest vs. abdominal radiographs, using only 45 training cases. For more difficult datasets, including the presence/absence and low/normal position endotracheal tubes, more training cases, pre-trained networks, and data-augmentation approaches were helpful to increase accuracy. The best-performing network for classifying presence vs. absence of an ET tube was still very accurate with an AUC of 0.99. However, for the most difficult dataset, such as low vs. normal position of the endotracheal tube, DCNNs did not perform as well, but achieved a reasonable AUC of 0.81.

The Hubble Deep UV Legacy Survey (HDUV): Survey Overview and First Results

NASA Astrophysics Data System (ADS)

Oesch, Pascal; Montes, Mireia; HDUV Survey Team

2015-08-01

Deep HST imaging has shown that the overall star formation density and UV light density at z>3 is dominated by faint, blue galaxies. Remarkably, very little is known about the equivalent galaxy population at lower redshifts. Understanding how these galaxies evolve across the epoch of peak cosmic star-formation is key to a complete picture of galaxy evolution. Here, we present a new HST WFC3/UVIS program, the Hubble Deep UV (HDUV) legacy survey. The HDUV is a 132 orbit program to obtain deep imaging in two filters (F275W and F336W) over the two CANDELS Deep fields. We will cover ~100 arcmin2, reaching down to 27.5-28.0 mag at 5 sigma. By directly sampling the rest-frame far-UV at z>~0.5, this will provide a unique legacy dataset with exquisite HST multi-wavelength imaging as well as ancillary HST grism NIR spectroscopy for a detailed study of faint, star-forming galaxies at z~0.5-2. The HDUV will enable a wealth of research by the community, which includes tracing the evolution of the FUV luminosity function over the peak of the star formation rate density from z~3 down to z~0.5, measuring the physical properties of sub-L* galaxies, and characterizing resolved stellar populations to decipher the build-up of the Hubble sequence from sub-galactic clumps. This poster provides an overview of the HDUV survey and presents the reduced data products and catalogs which will be released to the community.

"Whoa! We're Going Deep in the Trees!": Patterns of Collaboration around an Interactive Information Visualization Exhibit

ERIC Educational Resources Information Center

Davis, Pryce; Horn, Michael; Block, Florian; Phillips, Brenda; Evans, E. Margaret; Diamond, Judy; Shen, Chia

2015-01-01

In this paper we present a qualitative analysis of natural history museum visitor interaction around a multi-touch tabletop exhibit called "DeepTree" that we designed around concepts of evolution and common descent. DeepTree combines several large scientific datasets and an innovative visualization technique to display a phylogenetic…

Scientific drilling projects in ancient lakes: Integrating geological and biological histories

NASA Astrophysics Data System (ADS)

Wilke, Thomas; Wagner, Bernd; Van Bocxlaer, Bert; Albrecht, Christian; Ariztegui, Daniel; Delicado, Diana; Francke, Alexander; Harzhauser, Mathias; Hauffe, Torsten; Holtvoeth, Jens; Just, Janna; Leng, Melanie J.; Levkov, Zlatko; Penkman, Kirsty; Sadori, Laura; Skinner, Alister; Stelbrink, Björn; Vogel, Hendrik; Wesselingh, Frank; Wonik, Thomas

2016-08-01

Sedimentary sequences in ancient or long-lived lakes can reach several thousands of meters in thickness and often provide an unrivalled perspective of the lake's regional climatic, environmental, and biological history. Over the last few years, deep-drilling projects in ancient lakes became increasingly multi- and interdisciplinary, as, among others, seismological, sedimentological, biogeochemical, climatic, environmental, paleontological, and evolutionary information can be obtained from sediment cores. However, these multi- and interdisciplinary projects pose several challenges. The scientists involved typically approach problems from different scientific perspectives and backgrounds, and setting up the program requires clear communication and the alignment of interests. One of the most challenging tasks, besides the actual drilling operation, is to link diverse datasets with varying resolution, data quality, and age uncertainties to answer interdisciplinary questions synthetically and coherently. These problems are especially relevant when secondary data, i.e., datasets obtained independently of the drilling operation, are incorporated in analyses. Nonetheless, the inclusion of secondary information, such as isotopic data from fossils found in outcrops or genetic data from extant species, may help to achieve synthetic answers. Recent technological and methodological advances in paleolimnology are likely to increase the possibilities of integrating secondary information. Some of the new approaches have started to revolutionize scientific drilling in ancient lakes, but at the same time, they also add a new layer of complexity to the generation and analysis of sediment-core data. The enhanced opportunities presented by new scientific approaches to study the paleolimnological history of these lakes, therefore, come at the expense of higher logistic, communication, and analytical efforts. Here we review types of data that can be obtained in ancient lake drilling projects and the analytical approaches that can be applied to empirically and statistically link diverse datasets to create an integrative perspective on geological and biological data. In doing so, we highlight strengths and potential weaknesses of new methods and analyses, and provide recommendations for future interdisciplinary deep-drilling projects.

Biogeographic patterns of bacterial microdiversity in Arctic deep-sea sediments (HAUSGARTEN, Fram Strait)

PubMed Central

Buttigieg, Pier Luigi; Ramette, Alban

2015-01-01

Marine bacteria colonizing deep-sea sediments beneath the Arctic ocean, a rapidly changing ecosystem, have been shown to exhibit significant biogeographic patterns along transects spanning tens of kilometers and across water depths of several thousand meters (Jacob et al., 2013). Jacob et al. (2013) adopted what has become a classical view of microbial diversity – based on operational taxonomic units clustered at the 97% sequence identity level of the 16S rRNA gene – and observed a very large microbial community replacement at the HAUSGARTEN Long Term Ecological Research station (Eastern Fram Strait). Here, we revisited these data using the oligotyping approach and aimed to obtain new insight into ecological and biogeographic patterns associated with bacterial microdiversity in marine sediments. We also assessed the level of concordance of these insights with previously obtained results. Variation in oligotype dispersal range, relative abundance, co-occurrence, and taxonomic identity were related to environmental parameters such as water depth, biomass, and sedimentary pigment concentration. This study assesses ecological implications of the new microdiversity-based technique using a well-characterized dataset of high relevance for global change biology. PMID:25601856

OCT-based deep learning algorithm for the evaluation of treatment indication with anti-vascular endothelial growth factor medications.

PubMed

Prahs, Philipp; Radeck, Viola; Mayer, Christian; Cvetkov, Yordan; Cvetkova, Nadezhda; Helbig, Horst; Märker, David

2018-01-01

Intravitreal injections with anti-vascular endothelial growth factor (anti-VEGF) medications have become the standard of care for their respective indications. Optical coherence tomography (OCT) scans of the central retina provide detailed anatomical data and are widely used by clinicians in the decision-making process of anti-VEGF indication. In recent years, significant progress has been made in artificial intelligence and computer vision research. We trained a deep convolutional artificial neural network to predict treatment indication based on central retinal OCT scans without human intervention. A total of 183,402 retinal OCT B-scans acquired between 2008 and 2016 were exported from the institutional image archive of a university hospital. OCT images were cross-referenced with the electronic institutional intravitreal injection records. OCT images with a following intravitreal injection during the first 21 days after image acquisition were assigned into the 'injection' group, while the same amount of random OCT images without intravitreal injections was labeled as 'no injection'. After image preprocessing, OCT images were split in a 9:1 ratio to training and test datasets. We trained a GoogLeNet inception deep convolutional neural network and assessed its performance on the validation dataset. We calculated prediction accuracy, sensitivity, specificity, and receiver operating characteristics. The deep convolutional neural network was successfully trained on the extracted clinical data. The trained neural network classifier reached a prediction accuracy of 95.5% on the images in the validation dataset. For single retinal B-scans in the validation dataset, a sensitivity of 90.1% and a specificity of 96.2% were achieved. The area under the receiver operating characteristic curve was 0.968 on a per B-scan image basis, and 0.988 by averaging over six B-scans per examination on the validation dataset. Deep artificial neural networks show impressive performance on classification of retinal OCT scans. After training on historical clinical data, machine learning methods can offer the clinician support in the decision-making process. Care should be taken not to mistake neural network output as treatment recommendation and to ensure a final thorough evaluation by the treating physician.

Biophysics of protein evolution and evolutionary protein biophysics

PubMed Central

Sikosek, Tobias; Chan, Hue Sun

2014-01-01

The study of molecular evolution at the level of protein-coding genes often entails comparing large datasets of sequences to infer their evolutionary relationships. Despite the importance of a protein's structure and conformational dynamics to its function and thus its fitness, common phylogenetic methods embody minimal biophysical knowledge of proteins. To underscore the biophysical constraints on natural selection, we survey effects of protein mutations, highlighting the physical basis for marginal stability of natural globular proteins and how requirement for kinetic stability and avoidance of misfolding and misinteractions might have affected protein evolution. The biophysical underpinnings of these effects have been addressed by models with an explicit coarse-grained spatial representation of the polypeptide chain. Sequence–structure mappings based on such models are powerful conceptual tools that rationalize mutational robustness, evolvability, epistasis, promiscuous function performed by ‘hidden’ conformational states, resolution of adaptive conflicts and conformational switches in the evolution from one protein fold to another. Recently, protein biophysics has been applied to derive more accurate evolutionary accounts of sequence data. Methods have also been developed to exploit sequence-based evolutionary information to predict biophysical behaviours of proteins. The success of these approaches demonstrates a deep synergy between the fields of protein biophysics and protein evolution. PMID:25165599

Data Portal | Office of Cancer Clinical Proteomics Research

Cancer.gov

The CPTAC Data Portal is a centralized repository for the public dissemination of proteomic sequence datasets collected by CPTAC, along with corresponding genomic sequence datasets.  In addition, available are analyses of CPTAC's raw mass spectrometry-based data files (mapping of spectra to peptide sequences and protein identification) by individual investigators from CPTAC and by a Common Data Analysis Pipeline.

On the Multi-Modal Object Tracking and Image Fusion Using Unsupervised Deep Learning Methodologies

NASA Astrophysics Data System (ADS)

LaHaye, N.; Ott, J.; Garay, M. J.; El-Askary, H. M.; Linstead, E.

2017-12-01

The number of different modalities of remote-sensors has been on the rise, resulting in large datasets with different complexity levels. Such complex datasets can provide valuable information separately, yet there is a bigger value in having a comprehensive view of them combined. As such, hidden information can be deduced through applying data mining techniques on the fused data. The curse of dimensionality of such fused data, due to the potentially vast dimension space, hinders our ability to have deep understanding of them. This is because each dataset requires a user to have instrument-specific and dataset-specific knowledge for optimum and meaningful usage. Once a user decides to use multiple datasets together, deeper understanding of translating and combining these datasets in a correct and effective manner is needed. Although there exists data centric techniques, generic automated methodologies that can potentially solve this problem completely don't exist. Here we are developing a system that aims to gain a detailed understanding of different data modalities. Such system will provide an analysis environment that gives the user useful feedback and can aid in research tasks. In our current work, we show the initial outputs our system implementation that leverages unsupervised deep learning techniques so not to burden the user with the task of labeling input data, while still allowing for a detailed machine understanding of the data. Our goal is to be able to track objects, like cloud systems or aerosols, across different image-like data-modalities. The proposed system is flexible, scalable and robust to understand complex likenesses within multi-modal data in a similar spatio-temporal range, and also to be able to co-register and fuse these images when needed.

Transcriptome deep-sequencing and clustering of expressed isoforms from Favia corals

PubMed Central

2013-01-01

Background Genomic and transcriptomic sequence data are essential tools for tackling ecological problems. Using an approach that combines next-generation sequencing, de novo transcriptome assembly, gene annotation and synthetic gene construction, we identify and cluster the protein families from Favia corals from the northern Red Sea. Results We obtained 80 million 75 bp paired-end cDNA reads from two Favia adult samples collected at 65 m (Fav1, Fav2) on the Illumina GA platform, and generated two de novo assemblies using ABySS and CAP3. After removing redundancy and filtering out low quality reads, our transcriptome datasets contained 58,268 (Fav1) and 62,469 (Fav2) contigs longer than 100 bp, with N50 values of 1,665 bp and 1,439 bp, respectively. Using the proteome of the sea anemone Nematostella vectensis as a reference, we were able to annotate almost 20% of each dataset using reciprocal homology searches. Homologous clustering of these annotated transcripts allowed us to divide them into 7,186 (Fav1) and 6,862 (Fav2) homologous transcript clusters (E-value ≤ 2e-30). Functional annotation categories were assigned to homologous clusters using the functional annotation of Nematostella vectensis. General annotation of the assembled transcripts was improved 1-3% using the Acropora digitifera proteome. In addition, we screened these transcript isoform clusters for fluorescent proteins (FPs) homologs and identified seven potential FP homologs in Fav1, and four in Fav2. These transcripts were validated as bona fide FP transcripts via robust fluorescence heterologous expression. Annotation of the assembled contigs revealed that 1.34% and 1.61% (in Fav1 and Fav2, respectively) of the total assembled contigs likely originated from the corals’ algal symbiont, Symbiodinium spp. Conclusions Here we present a study to identify the homologous transcript isoform clusters from the transcriptome of Favia corals using a far-related reference proteome. Furthermore, the symbiont-derived transcripts were isolated from the datasets and their contribution quantified. This is the first annotated transcriptome of the genus Favia, a major increase in genomics resources available in this important family of corals. PMID:23937070

A Hybrid Spectral Clustering and Deep Neural Network Ensemble Algorithm for Intrusion Detection in Sensor Networks

PubMed Central

Ma, Tao; Wang, Fen; Cheng, Jianjun; Yu, Yang; Chen, Xiaoyun

2016-01-01

The development of intrusion detection systems (IDS) that are adapted to allow routers and network defence systems to detect malicious network traffic disguised as network protocols or normal access is a critical challenge. This paper proposes a novel approach called SCDNN, which combines spectral clustering (SC) and deep neural network (DNN) algorithms. First, the dataset is divided into k subsets based on sample similarity using cluster centres, as in SC. Next, the distance between data points in a testing set and the training set is measured based on similarity features and is fed into the deep neural network algorithm for intrusion detection. Six KDD-Cup99 and NSL-KDD datasets and a sensor network dataset were employed to test the performance of the model. These experimental results indicate that the SCDNN classifier not only performs better than backpropagation neural network (BPNN), support vector machine (SVM), random forest (RF) and Bayes tree models in detection accuracy and the types of abnormal attacks found. It also provides an effective tool of study and analysis of intrusion detection in large networks. PMID:27754380

A Hybrid Spectral Clustering and Deep Neural Network Ensemble Algorithm for Intrusion Detection in Sensor Networks.

PubMed

Ma, Tao; Wang, Fen; Cheng, Jianjun; Yu, Yang; Chen, Xiaoyun

2016-10-13

The development of intrusion detection systems (IDS) that are adapted to allow routers and network defence systems to detect malicious network traffic disguised as network protocols or normal access is a critical challenge. This paper proposes a novel approach called SCDNN, which combines spectral clustering (SC) and deep neural network (DNN) algorithms. First, the dataset is divided into k subsets based on sample similarity using cluster centres, as in SC. Next, the distance between data points in a testing set and the training set is measured based on similarity features and is fed into the deep neural network algorithm for intrusion detection. Six KDD-Cup99 and NSL-KDD datasets and a sensor network dataset were employed to test the performance of the model. These experimental results indicate that the SCDNN classifier not only performs better than backpropagation neural network (BPNN), support vector machine (SVM), random forest (RF) and Bayes tree models in detection accuracy and the types of abnormal attacks found. It also provides an effective tool of study and analysis of intrusion detection in large networks.

SPAR: small RNA-seq portal for analysis of sequencing experiments.

PubMed

Kuksa, Pavel P; Amlie-Wolf, Alexandre; Katanic, Živadin; Valladares, Otto; Wang, Li-San; Leung, Yuk Yee

2018-05-04

The introduction of new high-throughput small RNA sequencing protocols that generate large-scale genomics datasets along with increasing evidence of the significant regulatory roles of small non-coding RNAs (sncRNAs) have highlighted the urgent need for tools to analyze and interpret large amounts of small RNA sequencing data. However, it remains challenging to systematically and comprehensively discover and characterize sncRNA genes and specifically-processed sncRNA products from these datasets. To fill this gap, we present Small RNA-seq Portal for Analysis of sequencing expeRiments (SPAR), a user-friendly web server for interactive processing, analysis, annotation and visualization of small RNA sequencing data. SPAR supports sequencing data generated from various experimental protocols, including smRNA-seq, short total RNA sequencing, microRNA-seq, and single-cell small RNA-seq. Additionally, SPAR includes publicly available reference sncRNA datasets from our DASHR database and from ENCODE across 185 human tissues and cell types to produce highly informative small RNA annotations across all major small RNA types and other features such as co-localization with various genomic features, precursor transcript cleavage patterns, and conservation. SPAR allows the user to compare the input experiment against reference ENCODE/DASHR datasets. SPAR currently supports analyses of human (hg19, hg38) and mouse (mm10) sequencing data. SPAR is freely available at https://www.lisanwanglab.org/SPAR.

An efficient and scalable graph modeling approach for capturing information at different levels in next generation sequencing reads

PubMed Central

2013-01-01

Background Next generation sequencing technologies have greatly advanced many research areas of the biomedical sciences through their capability to generate massive amounts of genetic information at unprecedented rates. The advent of next generation sequencing has led to the development of numerous computational tools to analyze and assemble the millions to billions of short sequencing reads produced by these technologies. While these tools filled an important gap, current approaches for storing, processing, and analyzing short read datasets generally have remained simple and lack the complexity needed to efficiently model the produced reads and assemble them correctly. Results Previously, we presented an overlap graph coarsening scheme for modeling read overlap relationships on multiple levels. Most current read assembly and analysis approaches use a single graph or set of clusters to represent the relationships among a read dataset. Instead, we use a series of graphs to represent the reads and their overlap relationships across a spectrum of information granularity. At each information level our algorithm is capable of generating clusters of reads from the reduced graph, forming an integrated graph modeling and clustering approach for read analysis and assembly. Previously we applied our algorithm to simulated and real 454 datasets to assess its ability to efficiently model and cluster next generation sequencing data. In this paper we extend our algorithm to large simulated and real Illumina datasets to demonstrate that our algorithm is practical for both sequencing technologies. Conclusions Our overlap graph theoretic algorithm is able to model next generation sequencing reads at various levels of granularity through the process of graph coarsening. Additionally, our model allows for efficient representation of the read overlap relationships, is scalable for large datasets, and is practical for both Illumina and 454 sequencing technologies. PMID:24564333

An analysis of geothermal and carbonic springs in the western United States sustained by deep fluid inputs.

PubMed

Colman, D R; Garcia, J R; Crossey, L J; Karlstrom, K; Jackson-Weaver, O; Takacs-Vesbach, C

2014-01-01

Hydrothermal springs harbor unique microbial communities that have provided insight into the early evolution of life, expanded known microbial diversity, and documented a deep Earth biosphere. Mesothermal (cool but above ambient temperature) continental springs, however, have largely been ignored although they may also harbor unique populations of micro-organisms influenced by deep subsurface fluid mixing with near surface fluids. We investigated the microbial communities of 28 mesothermal springs in diverse geologic provinces of the western United States that demonstrate differential mixing of deeply and shallowly circulated water. Culture-independent analysis of the communities yielded 1966 bacterial and 283 archaeal 16S rRNA gene sequences. The springs harbored diverse taxa and shared few operational taxonomic units (OTUs) across sites. The Proteobacteria phylum accounted for most of the dataset (81.2% of all 16S rRNA genes), with 31 other phyla/candidate divisions comprising the remainder. A small percentage (~6%) of bacterial 16S rRNA genes could not be classified at the phylum level, but were mostly distributed in those springs with greatest inputs of deeply sourced fluids. Archaeal diversity was limited to only four springs and was primarily composed of well-characterized Thaumarchaeota. Geochemistry across the dataset was varied, but statistical analyses suggested that greater input of deeply sourced fluids was correlated with community structure. Those with lesser input contained genera typical of surficial waters, while some of the springs with greater input may contain putatively chemolithotrophic communities. The results reported here expand our understanding of microbial diversity of continental geothermal systems and suggest that these communities are influenced by the geochemical and hydrologic characteristics arising from deeply sourced (mantle-derived) fluid mixing. The springs and communities we report here provide evidence for opportunities to understand new dimensions of continental geobiological processes where warm, highly reduced fluids are mixing with more oxidized surficial waters. © 2013 John Wiley & Sons Ltd.

A computational method for estimating the PCR duplication rate in DNA and RNA-seq experiments.

PubMed

Bansal, Vikas

2017-03-14

PCR amplification is an important step in the preparation of DNA sequencing libraries prior to high-throughput sequencing. PCR amplification introduces redundant reads in the sequence data and estimating the PCR duplication rate is important to assess the frequency of such reads. Existing computational methods do not distinguish PCR duplicates from "natural" read duplicates that represent independent DNA fragments and therefore, over-estimate the PCR duplication rate for DNA-seq and RNA-seq experiments. In this paper, we present a computational method to estimate the average PCR duplication rate of high-throughput sequence datasets that accounts for natural read duplicates by leveraging heterozygous variants in an individual genome. Analysis of simulated data and exome sequence data from the 1000 Genomes project demonstrated that our method can accurately estimate the PCR duplication rate on paired-end as well as single-end read datasets which contain a high proportion of natural read duplicates. Further, analysis of exome datasets prepared using the Nextera library preparation method indicated that 45-50% of read duplicates correspond to natural read duplicates likely due to fragmentation bias. Finally, analysis of RNA-seq datasets from individuals in the 1000 Genomes project demonstrated that 70-95% of read duplicates observed in such datasets correspond to natural duplicates sampled from genes with high expression and identified outlier samples with a 2-fold greater PCR duplication rate than other samples. The method described here is a useful tool for estimating the PCR duplication rate of high-throughput sequence datasets and for assessing the fraction of read duplicates that correspond to natural read duplicates. An implementation of the method is available at https://github.com/vibansal/PCRduplicates .

Insights into Deep-Sea Sediment Fungal Communities from the East Indian Ocean Using Targeted Environmental Sequencing Combined with Traditional Cultivation

PubMed Central

Zhang, Xiao-yong; Tang, Gui-ling; Xu, Xin-ya; Nong, Xu-hua; Qi, Shu-Hua

2014-01-01

The fungal diversity in deep-sea environments has recently gained an increasing amount attention. Our knowledge and understanding of the true fungal diversity and the role it plays in deep-sea environments, however, is still limited. We investigated the fungal community structure in five sediments from a depth of ∼4000 m in the East India Ocean using a combination of targeted environmental sequencing and traditional cultivation. This approach resulted in the recovery of a total of 45 fungal operational taxonomic units (OTUs) and 20 culturable fungal phylotypes. This finding indicates that there is a great amount of fungal diversity in the deep-sea sediments collected in the East Indian Ocean. Three fungal OTUs and one culturable phylotype demonstrated high divergence (89%–97%) from the existing sequences in the GenBank. Moreover, 44.4% fungal OTUs and 30% culturable fungal phylotypes are new reports for deep-sea sediments. These results suggest that the deep-sea sediments from the East India Ocean can serve as habitats for new fungal communities compared with other deep-sea environments. In addition, different fungal community could be detected when using targeted environmental sequencing compared with traditional cultivation in this study, which suggests that a combination of targeted environmental sequencing and traditional cultivation will generate a more diverse fungal community in deep-sea environments than using either targeted environmental sequencing or traditional cultivation alone. This study is the first to report new insights into the fungal communities in deep-sea sediments from the East Indian Ocean, which increases our knowledge and understanding of the fungal diversity in deep-sea environments. PMID:25272044

Deep learning-based fine-grained car make/model classification for visual surveillance

NASA Astrophysics Data System (ADS)

Gundogdu, Erhan; Parıldı, Enes Sinan; Solmaz, Berkan; Yücesoy, Veysel; Koç, Aykut

2017-10-01

Fine-grained object recognition is a potential computer vision problem that has been recently addressed by utilizing deep Convolutional Neural Networks (CNNs). Nevertheless, the main disadvantage of classification methods relying on deep CNN models is the need for considerably large amount of data. In addition, there exists relatively less amount of annotated data for a real world application, such as the recognition of car models in a traffic surveillance system. To this end, we mainly concentrate on the classification of fine-grained car make and/or models for visual scenarios by the help of two different domains. First, a large-scale dataset including approximately 900K images is constructed from a website which includes fine-grained car models. According to their labels, a state-of-the-art CNN model is trained on the constructed dataset. The second domain that is dealt with is the set of images collected from a camera integrated to a traffic surveillance system. These images, which are over 260K, are gathered by a special license plate detection method on top of a motion detection algorithm. An appropriately selected size of the image is cropped from the region of interest provided by the detected license plate location. These sets of images and their provided labels for more than 30 classes are employed to fine-tune the CNN model which is already trained on the large scale dataset described above. To fine-tune the network, the last two fully-connected layers are randomly initialized and the remaining layers are fine-tuned in the second dataset. In this work, the transfer of a learned model on a large dataset to a smaller one has been successfully performed by utilizing both the limited annotated data of the traffic field and a large scale dataset with available annotations. Our experimental results both in the validation dataset and the real field show that the proposed methodology performs favorably against the training of the CNN model from scratch.

Discrimination of Breast Cancer with Microcalcifications on Mammography by Deep Learning.

PubMed

Wang, Jinhua; Yang, Xi; Cai, Hongmin; Tan, Wanchang; Jin, Cangzheng; Li, Li

2016-06-07

Microcalcification is an effective indicator of early breast cancer. To improve the diagnostic accuracy of microcalcifications, this study evaluates the performance of deep learning-based models on large datasets for its discrimination. A semi-automated segmentation method was used to characterize all microcalcifications. A discrimination classifier model was constructed to assess the accuracies of microcalcifications and breast masses, either in isolation or combination, for classifying breast lesions. Performances were compared to benchmark models. Our deep learning model achieved a discriminative accuracy of 87.3% if microcalcifications were characterized alone, compared to 85.8% with a support vector machine. The accuracies were 61.3% for both methods with masses alone and improved to 89.7% and 85.8% after the combined analysis with microcalcifications. Image segmentation with our deep learning model yielded 15, 26 and 41 features for the three scenarios, respectively. Overall, deep learning based on large datasets was superior to standard methods for the discrimination of microcalcifications. Accuracy was increased by adopting a combinatorial approach to detect microcalcifications and masses simultaneously. This may have clinical value for early detection and treatment of breast cancer.

Discrimination of Breast Cancer with Microcalcifications on Mammography by Deep Learning

PubMed Central

Wang, Jinhua; Yang, Xi; Cai, Hongmin; Tan, Wanchang; Jin, Cangzheng; Li, Li

2016-01-01

Microcalcification is an effective indicator of early breast cancer. To improve the diagnostic accuracy of microcalcifications, this study evaluates the performance of deep learning-based models on large datasets for its discrimination. A semi-automated segmentation method was used to characterize all microcalcifications. A discrimination classifier model was constructed to assess the accuracies of microcalcifications and breast masses, either in isolation or combination, for classifying breast lesions. Performances were compared to benchmark models. Our deep learning model achieved a discriminative accuracy of 87.3% if microcalcifications were characterized alone, compared to 85.8% with a support vector machine. The accuracies were 61.3% for both methods with masses alone and improved to 89.7% and 85.8% after the combined analysis with microcalcifications. Image segmentation with our deep learning model yielded 15, 26 and 41 features for the three scenarios, respectively. Overall, deep learning based on large datasets was superior to standard methods for the discrimination of microcalcifications. Accuracy was increased by adopting a combinatorial approach to detect microcalcifications and masses simultaneously. This may have clinical value for early detection and treatment of breast cancer. PMID:27273294

Predicting Virtual World User Population Fluctuations with Deep Learning

PubMed Central

Park, Nuri; Zhang, Qimeng; Kim, Jun Gi; Kang, Shin Jin; Kim, Chang Hun

2016-01-01

This paper proposes a system for predicting increases in virtual world user actions. The virtual world user population is a very important aspect of these worlds; however, methods for predicting fluctuations in these populations have not been well documented. Therefore, we attempt to predict changes in virtual world user populations with deep learning, using easily accessible online data, including formal datasets from Google Trends, Wikipedia, and online communities, as well as informal datasets collected from online forums. We use the proposed system to analyze the user population of EVE Online, one of the largest virtual worlds. PMID:27936009

Predicting Virtual World User Population Fluctuations with Deep Learning.

PubMed

Kim, Young Bin; Park, Nuri; Zhang, Qimeng; Kim, Jun Gi; Kang, Shin Jin; Kim, Chang Hun

2016-01-01

This paper proposes a system for predicting increases in virtual world user actions. The virtual world user population is a very important aspect of these worlds; however, methods for predicting fluctuations in these populations have not been well documented. Therefore, we attempt to predict changes in virtual world user populations with deep learning, using easily accessible online data, including formal datasets from Google Trends, Wikipedia, and online communities, as well as informal datasets collected from online forums. We use the proposed system to analyze the user population of EVE Online, one of the largest virtual worlds.

ESTuber db: an online database for Tuber borchii EST sequences.

PubMed

Lazzari, Barbara; Caprera, Andrea; Cosentino, Cristian; Stella, Alessandra; Milanesi, Luciano; Viotti, Angelo

2007-03-08

The ESTuber database (http://www.itb.cnr.it/estuber) includes 3,271 Tuber borchii expressed sequence tags (EST). The dataset consists of 2,389 sequences from an in-house prepared cDNA library from truffle vegetative hyphae, and 882 sequences downloaded from GenBank and representing four libraries from white truffle mycelia and ascocarps at different developmental stages. An automated pipeline was prepared to process EST sequences using public software integrated by in-house developed Perl scripts. Data were collected in a MySQL database, which can be queried via a php-based web interface. Sequences included in the ESTuber db were clustered and annotated against three databases: the GenBank nr database, the UniProtKB database and a third in-house prepared database of fungi genomic sequences. An algorithm was implemented to infer statistical classification among Gene Ontology categories from the ontology occurrences deduced from the annotation procedure against the UniProtKB database. Ontologies were also deduced from the annotation of more than 130,000 EST sequences from five filamentous fungi, for intra-species comparison purposes. Further analyses were performed on the ESTuber db dataset, including tandem repeats search and comparison of the putative protein dataset inferred from the EST sequences to the PROSITE database for protein patterns identification. All the analyses were performed both on the complete sequence dataset and on the contig consensus sequences generated by the EST assembly procedure. The resulting web site is a resource of data and links related to truffle expressed genes. The Sequence Report and Contig Report pages are the web interface core structures which, together with the Text search utility and the Blast utility, allow easy access to the data stored in the database.

Comparing MODIS C6 'Deep Blue' and 'Dark Target' Aerosol Data

NASA Technical Reports Server (NTRS)

Hsu, N. C.; Sayer, A. M.; Bettenhausen, C.; Lee, J.; Levy, R. C.; Mattoo, S.; Munchak, L. A.; Kleidman, R.

2014-01-01

The MODIS Collection 6 Atmospheres product suite includes refined versions of both 'Deep Blue' (DB) and 'Dark Target' (DT) aerosol algorithms, with the DB dataset now expanded to include coverage over vegetated land surfaces. This means that, over much of the global land surface, users will have both DB and DT data to choose from. A 'merged' dataset is also provided, primarily for visualization purposes, which takes retrievals from either or both algorithms based on regional and seasonal climatologies of normalized difference vegetation index (NDVI). This poster present some comparisons of these two C6 aerosol algorithms, focusing on AOD at 550 nm derived from MODIS Aqua measurements, with each other and with Aerosol Robotic Network (AERONET) data, with the intent to facilitate user decisions about the suitability of the two datasets for their desired applications.

TriageTools: tools for partitioning and prioritizing analysis of high-throughput sequencing data.

PubMed

Fimereli, Danai; Detours, Vincent; Konopka, Tomasz

2013-04-01

High-throughput sequencing is becoming a popular research tool but carries with it considerable costs in terms of computation time, data storage and bandwidth. Meanwhile, some research applications focusing on individual genes or pathways do not necessitate processing of a full sequencing dataset. Thus, it is desirable to partition a large dataset into smaller, manageable, but relevant pieces. We present a toolkit for partitioning raw sequencing data that includes a method for extracting reads that are likely to map onto pre-defined regions of interest. We show the method can be used to extract information about genes of interest from DNA or RNA sequencing samples in a fraction of the time and disk space required to process and store a full dataset. We report speedup factors between 2.6 and 96, depending on settings and samples used. The software is available at http://www.sourceforge.net/projects/triagetools/.

PASTA for Proteins.

PubMed

Collins, Kodi; Warnow, Tandy

2018-06-19

PASTA is a multiple sequence method that uses divide-and-conquer plus iteration to enable base alignment methods to scale with high accuracy to large sequence datasets. By default, PASTA included MAFFT L-INS-i; our new extension of PASTA enables the use of MAFFT G-INS-i, MAFFT Homologs, CONTRAlign, and ProbCons. We analyzed the performance of each base method and PASTA using these base methods on 224 datasets from BAliBASE 4 with at least 50 sequences. We show that PASTA enables the most accurate base methods to scale to larger datasets at reduced computational effort, and generally improves alignment and tree accuracy on the largest BAliBASE datasets. PASTA is available at https://github.com/kodicollins/pasta and has also been integrated into the original PASTA repository at https://github.com/smirarab/pasta. Supplementary data are available at Bioinformatics online.

Thermodynamic Data Rescue and Informatics for Deep Carbon Science

NASA Astrophysics Data System (ADS)

Zhong, H.; Ma, X.; Prabhu, A.; Eleish, A.; Pan, F.; Parsons, M. A.; Ghiorso, M. S.; West, P.; Zednik, S.; Erickson, J. S.; Chen, Y.; Wang, H.; Fox, P. A.

2017-12-01

A large number of legacy datasets are contained in geoscience literature published between 1930 and 1980 and not expressed external to the publication text in digitized formats. Extracting, organizing, and reusing these "dark" datasets is highly valuable for many within the Earth and planetary science community. As a part of the Deep Carbon Observatory (DCO) data legacy missions, the DCO Data Science Team and Extreme Physics and Chemistry community identified thermodynamic datasets related to carbon, or more specifically datasets about the enthalpy and entropy of chemicals, as a proof of principle analysis. The data science team endeavored to develop a semi-automatic workflow, which includes identifying relevant publications, extracting contained datasets using OCR methods, collaborative reviewing, and registering the datasets via the DCO Data Portal where the 'Linked Data' feature of the data portal provides a mechanism for connecting rescued datasets beyond their individual data sources, to research domains, DCO Communities, and more, making data discovery and retrieval more effective.To date, the team has successfully rescued, deposited and registered additional datasets from publications with thermodynamic sources. These datasets contain 3 main types of data: (1) heat content or enthalpy data determined for a given compound as a function of temperature using high-temperature calorimetry, (2) heat content or enthalpy data determined for a given compound as a function of temperature using adiabatic calorimetry, and (3) direct determination of heat capacity of a compound as a function of temperature using differential scanning calorimetry. The data science team integrated these datasets and delivered a spectrum of data analytics including visualizations, which will lead to a comprehensive characterization of the thermodynamics of carbon and carbon-related materials.

MIPE: A metagenome-based community structure explorer and SSU primer evaluation tool

PubMed Central

Zhou, Quan

2017-01-01

An understanding of microbial community structure is an important issue in the field of molecular ecology. The traditional molecular method involves amplification of small subunit ribosomal RNA (SSU rRNA) genes by polymerase chain reaction (PCR). However, PCR-based amplicon approaches are affected by primer bias and chimeras. With the development of high-throughput sequencing technology, unbiased SSU rRNA gene sequences can be mined from shotgun sequencing-based metagenomic or metatranscriptomic datasets to obtain a reflection of the microbial community structure in specific types of environment and to evaluate SSU primers. However, the use of short reads obtained through next-generation sequencing for primer evaluation has not been well resolved. The software MIPE (MIcrobiota metagenome Primer Explorer) was developed to adapt numerous short reads from metagenomes and metatranscriptomes. Using metagenomic or metatranscriptomic datasets as input, MIPE extracts and aligns rRNA to reveal detailed information on microbial composition and evaluate SSU rRNA primers. A mock dataset, a real Metagenomics Rapid Annotation using Subsystem Technology (MG-RAST) test dataset, two PrimerProspector test datasets and a real metatranscriptomic dataset were used to validate MIPE. The software calls Mothur (v1.33.3) and the SILVA database (v119) for the alignment and classification of rRNA genes from a metagenome or metatranscriptome. MIPE can effectively extract shotgun rRNA reads from a metagenome or metatranscriptome and is capable of classifying these sequences and exhibiting sensitivity to different SSU rRNA PCR primers. Therefore, MIPE can be used to guide primer design for specific environmental samples. PMID:28350876

Resolving Relationships among the Megadiverse Butterflies and Moths with a Novel Pipeline for Anchored Phylogenomics.

PubMed

Breinholt, Jesse W; Earl, Chandra; Lemmon, Alan R; Lemmon, Emily Moriarty; Xiao, Lei; Kawahara, Akito Y

2018-01-01

The advent of next-generation sequencing technology has allowed for thecollection of large portions of the genome for phylogenetic analysis. Hybrid enrichment and transcriptomics are two techniques that leverage next-generation sequencing and have shown much promise. However, methods for processing hybrid enrichment data are still limited. We developed a pipeline for anchored hybrid enrichment (AHE) read assembly, orthology determination, contamination screening, and data processing for sequences flanking the target "probe" region. We apply this approach to study the phylogeny of butterflies and moths (Lepidoptera), a megadiverse group of more than 157,000 described species with poorly understood deep-level phylogenetic relationships. We introduce a new, 855 locus AHE kit for Lepidoptera phylogenetics and compare resulting trees to those from transcriptomes. The enrichment kit was designed from existing genomes, transcriptomes, and expressed sequence tags and was used to capture sequence data from 54 species from 23 lepidopteran families. Phylogenies estimated from AHE data were largely congruent with trees generated from transcriptomes, with strong support for relationships at all but the deepest taxonomic levels. We combine AHE and transcriptomic data to generate a new Lepidoptera phylogeny, representing 76 exemplar species in 42 families. The tree provides robust support for many relationships, including those among the seven butterfly families. The addition of AHE data to an existing transcriptomic dataset lowers node support along the Lepidoptera backbone, but firmly places taxa with AHE data on the phylogeny. Combining taxa sequenced for AHE with existing transcriptomes and genomes resulted in a tree with strong support for (Calliduloidea $+$ Gelechioidea $+$ Thyridoidea) $+$ (Papilionoidea $+$ Pyraloidea $+$ Macroheterocera). To examine the efficacy of AHE at a shallow taxonomic level, phylogenetic analyses were also conducted on a sister group representing a more recent divergence, the Saturniidae and Sphingidae. These analyses utilized sequences from the probe region and data flanking it, nearly doubled the size of the dataset; resulting trees supported new phylogenetics relationships, especially within the Saturniidae and Sphingidae (e.g., Hemarina derived in the latter). We hope that our data processing pipeline, hybrid enrichment gene set, and approach of combining AHE data with transcriptomes will be useful for the broader systematics community. © The Author(s) 2017. Published by Oxford University Press, on behalf of the Society of Systematic Biologists. All rights reserved. For Permissions, please email: journals.permissions@oup.com.

Quasispecies Analyses of the HIV-1 Near-full-length Genome With Illumina MiSeq

PubMed Central

Ode, Hirotaka; Matsuda, Masakazu; Matsuoka, Kazuhiro; Hachiya, Atsuko; Hattori, Junko; Kito, Yumiko; Yokomaku, Yoshiyuki; Iwatani, Yasumasa; Sugiura, Wataru

2015-01-01

Human immunodeficiency virus type-1 (HIV-1) exhibits high between-host genetic diversity and within-host heterogeneity, recognized as quasispecies. Because HIV-1 quasispecies fluctuate in terms of multiple factors, such as antiretroviral exposure and host immunity, analyzing the HIV-1 genome is critical for selecting effective antiretroviral therapy and understanding within-host viral coevolution mechanisms. Here, to obtain HIV-1 genome sequence information that includes minority variants, we sought to develop a method for evaluating quasispecies throughout the HIV-1 near-full-length genome using the Illumina MiSeq benchtop deep sequencer. To ensure the reliability of minority mutation detection, we applied an analysis method of sequence read mapping onto a consensus sequence derived from de novo assembly followed by iterative mapping and subsequent unique error correction. Deep sequencing analyses of aHIV-1 clone showed that the analysis method reduced erroneous base prevalence below 1% in each sequence position and discarded only < 1% of all collected nucleotides, maximizing the usage of the collected genome sequences. Further, we designed primer sets to amplify the HIV-1 near-full-length genome from clinical plasma samples. Deep sequencing of 92 samples in combination with the primer sets and our analysis method provided sufficient coverage to identify >1%-frequency sequences throughout the genome. When we evaluated sequences of pol genes from 18 treatment-naïve patients' samples, the deep sequencing results were in agreement with Sanger sequencing and identified numerous additional minority mutations. The results suggest that our deep sequencing method would be suitable for identifying within-host viral population dynamics throughout the genome. PMID:26617593

Convolutional networks for fast, energy-efficient neuromorphic computing

PubMed Central

Esser, Steven K.; Merolla, Paul A.; Arthur, John V.; Cassidy, Andrew S.; Appuswamy, Rathinakumar; Andreopoulos, Alexander; Berg, David J.; McKinstry, Jeffrey L.; Melano, Timothy; Barch, Davis R.; di Nolfo, Carmelo; Datta, Pallab; Amir, Arnon; Taba, Brian; Flickner, Myron D.; Modha, Dharmendra S.

2016-01-01

Deep networks are now able to achieve human-level performance on a broad spectrum of recognition tasks. Independently, neuromorphic computing has now demonstrated unprecedented energy-efficiency through a new chip architecture based on spiking neurons, low precision synapses, and a scalable communication network. Here, we demonstrate that neuromorphic computing, despite its novel architectural primitives, can implement deep convolution networks that (i) approach state-of-the-art classification accuracy across eight standard datasets encompassing vision and speech, (ii) perform inference while preserving the hardware’s underlying energy-efficiency and high throughput, running on the aforementioned datasets at between 1,200 and 2,600 frames/s and using between 25 and 275 mW (effectively >6,000 frames/s per Watt), and (iii) can be specified and trained using backpropagation with the same ease-of-use as contemporary deep learning. This approach allows the algorithmic power of deep learning to be merged with the efficiency of neuromorphic processors, bringing the promise of embedded, intelligent, brain-inspired computing one step closer. PMID:27651489

Convolutional networks for fast, energy-efficient neuromorphic computing.

PubMed

Esser, Steven K; Merolla, Paul A; Arthur, John V; Cassidy, Andrew S; Appuswamy, Rathinakumar; Andreopoulos, Alexander; Berg, David J; McKinstry, Jeffrey L; Melano, Timothy; Barch, Davis R; di Nolfo, Carmelo; Datta, Pallab; Amir, Arnon; Taba, Brian; Flickner, Myron D; Modha, Dharmendra S

2016-10-11

Deep networks are now able to achieve human-level performance on a broad spectrum of recognition tasks. Independently, neuromorphic computing has now demonstrated unprecedented energy-efficiency through a new chip architecture based on spiking neurons, low precision synapses, and a scalable communication network. Here, we demonstrate that neuromorphic computing, despite its novel architectural primitives, can implement deep convolution networks that (i) approach state-of-the-art classification accuracy across eight standard datasets encompassing vision and speech, (ii) perform inference while preserving the hardware's underlying energy-efficiency and high throughput, running on the aforementioned datasets at between 1,200 and 2,600 frames/s and using between 25 and 275 mW (effectively >6,000 frames/s per Watt), and (iii) can be specified and trained using backpropagation with the same ease-of-use as contemporary deep learning. This approach allows the algorithmic power of deep learning to be merged with the efficiency of neuromorphic processors, bringing the promise of embedded, intelligent, brain-inspired computing one step closer.

Virus Identification in Unknown Tropical Febrile Illness Cases Using Deep Sequencing

PubMed Central

Balmaseda, Angel; Harris, Eva; DeRisi, Joseph L.

2012-01-01

Dengue virus is an emerging infectious agent that infects an estimated 50–100 million people annually worldwide, yet current diagnostic practices cannot detect an etiologic pathogen in ∼40% of dengue-like illnesses. Metagenomic approaches to pathogen detection, such as viral microarrays and deep sequencing, are promising tools to address emerging and non-diagnosable disease challenges. In this study, we used the Virochip microarray and deep sequencing to characterize the spectrum of viruses present in human sera from 123 Nicaraguan patients presenting with dengue-like symptoms but testing negative for dengue virus. We utilized a barcoding strategy to simultaneously deep sequence multiple serum specimens, generating on average over 1 million reads per sample. We then implemented a stepwise bioinformatic filtering pipeline to remove the majority of human and low-quality sequences to improve the speed and accuracy of subsequent unbiased database searches. By deep sequencing, we were able to detect virus sequence in 37% (45/123) of previously negative cases. These included 13 cases with Human Herpesvirus 6 sequences. Other samples contained sequences with similarity to sequences from viruses in the Herpesviridae, Flaviviridae, Circoviridae, Anelloviridae, Asfarviridae, and Parvoviridae families. In some cases, the putative viral sequences were virtually identical to known viruses, and in others they diverged, suggesting that they may derive from novel viruses. These results demonstrate the utility of unbiased metagenomic approaches in the detection of known and divergent viruses in the study of tropical febrile illness. PMID:22347512

TAM: a method for enrichment and depletion analysis of a microRNA category in a list of microRNAs.

PubMed

Lu, Ming; Shi, Bing; Wang, Juan; Cao, Qun; Cui, Qinghua

2010-08-09

MicroRNAs (miRNAs) are a class of important gene regulators. The number of identified miRNAs has been increasing dramatically in recent years. An emerging major challenge is the interpretation of the genome-scale miRNA datasets, including those derived from microarray and deep-sequencing. It is interesting and important to know the common rules or patterns behind a list of miRNAs, (i.e. the deregulated miRNAs resulted from an experiment of miRNA microarray or deep-sequencing). For the above purpose, this study presents a method and develops a tool (TAM) for annotations of meaningful human miRNAs categories. We first integrated miRNAs into various meaningful categories according to prior knowledge, such as miRNA family, miRNA cluster, miRNA function, miRNA associated diseases, and tissue specificity. Using TAM, given lists of miRNAs can be rapidly annotated and summarized according to the integrated miRNA categorical data. Moreover, given a list of miRNAs, TAM can be used to predict novel related miRNAs. Finally, we confirmed the usefulness and reliability of TAM by applying it to deregulated miRNAs in acute myocardial infarction (AMI) from two independent experiments. TAM can efficiently identify meaningful categories for given miRNAs. In addition, TAM can be used to identify novel miRNA biomarkers. TAM tool, source codes, and miRNA category data are freely available at http://cmbi.bjmu.edu.cn/tam.

A novel end-to-end classifier using domain transferred deep convolutional neural networks for biomedical images.

PubMed

Pang, Shuchao; Yu, Zhezhou; Orgun, Mehmet A

2017-03-01

Highly accurate classification of biomedical images is an essential task in the clinical diagnosis of numerous medical diseases identified from those images. Traditional image classification methods combined with hand-crafted image feature descriptors and various classifiers are not able to effectively improve the accuracy rate and meet the high requirements of classification of biomedical images. The same also holds true for artificial neural network models directly trained with limited biomedical images used as training data or directly used as a black box to extract the deep features based on another distant dataset. In this study, we propose a highly reliable and accurate end-to-end classifier for all kinds of biomedical images via deep learning and transfer learning. We first apply domain transferred deep convolutional neural network for building a deep model; and then develop an overall deep learning architecture based on the raw pixels of original biomedical images using supervised training. In our model, we do not need the manual design of the feature space, seek an effective feature vector classifier or segment specific detection object and image patches, which are the main technological difficulties in the adoption of traditional image classification methods. Moreover, we do not need to be concerned with whether there are large training sets of annotated biomedical images, affordable parallel computing resources featuring GPUs or long times to wait for training a perfect deep model, which are the main problems to train deep neural networks for biomedical image classification as observed in recent works. With the utilization of a simple data augmentation method and fast convergence speed, our algorithm can achieve the best accuracy rate and outstanding classification ability for biomedical images. We have evaluated our classifier on several well-known public biomedical datasets and compared it with several state-of-the-art approaches. We propose a robust automated end-to-end classifier for biomedical images based on a domain transferred deep convolutional neural network model that shows a highly reliable and accurate performance which has been confirmed on several public biomedical image datasets. Copyright © 2017 Elsevier Ireland Ltd. All rights reserved.

Leveraging genome-wide datasets to quantify the functional role of the anti-Shine-Dalgarno sequence in regulating translation efficiency.

PubMed

Hockenberry, Adam J; Pah, Adam R; Jewett, Michael C; Amaral, Luís A N

2017-01-01

Studies dating back to the 1970s established that sequence complementarity between the anti-Shine-Dalgarno (aSD) sequence on prokaryotic ribosomes and the 5' untranslated region of mRNAs helps to facilitate translation initiation. The optimal location of aSD sequence binding relative to the start codon, the full extents of the aSD sequence and the functional form of the relationship between aSD sequence complementarity and translation efficiency have not been fully resolved. Here, we investigate these relationships by leveraging the sequence diversity of endogenous genes and recently available genome-wide estimates of translation efficiency. We show that-after accounting for predicted mRNA structure-aSD sequence complementarity increases the translation of endogenous mRNAs by roughly 50%. Further, we observe that this relationship is nonlinear, with translation efficiency maximized for mRNAs with intermediate levels of aSD sequence complementarity. The mechanistic insights that we observe are highly robust: we find nearly identical results in multiple datasets spanning three distantly related bacteria. Further, we verify our main conclusions by re-analysing a controlled experimental dataset. © 2017 The Authors.

Major soybean maturity gene haplotypes revealed by SNPViz analysis of 72 sequenced soybean genomes

USDA-ARS?s Scientific Manuscript database

In this Genomics Era, vast amounts of next generation sequencing data have become publicly-available for multiple genomes across hundreds of species. Analysis of these large-scale datasets can become cumbersome, especially when comparing nucleotide polymorphisms across many samples within a dataset...

deepTools: a flexible platform for exploring deep-sequencing data.

PubMed

Ramírez, Fidel; Dündar, Friederike; Diehl, Sarah; Grüning, Björn A; Manke, Thomas

2014-07-01

We present a Galaxy based web server for processing and visualizing deeply sequenced data. The web server's core functionality consists of a suite of newly developed tools, called deepTools, that enable users with little bioinformatic background to explore the results of their sequencing experiments in a standardized setting. Users can upload pre-processed files with continuous data in standard formats and generate heatmaps and summary plots in a straight-forward, yet highly customizable manner. In addition, we offer several tools for the analysis of files containing aligned reads and enable efficient and reproducible generation of normalized coverage files. As a modular and open-source platform, deepTools can easily be expanded and customized to future demands and developments. The deepTools webserver is freely available at http://deeptools.ie-freiburg.mpg.de and is accompanied by extensive documentation and tutorials aimed at conveying the principles of deep-sequencing data analysis. The web server can be used without registration. deepTools can be installed locally either stand-alone or as part of Galaxy. © The Author(s) 2014. Published by Oxford University Press on behalf of Nucleic Acids Research.

Computer-aided classification of mammographic masses using the deep learning technology: a preliminary study

NASA Astrophysics Data System (ADS)

Qiu, Yuchen; Yan, Shiju; Tan, Maxine; Cheng, Samuel; Liu, Hong; Zheng, Bin

2016-03-01

Although mammography is the only clinically acceptable imaging modality used in the population-based breast cancer screening, its efficacy is quite controversy. One of the major challenges is how to help radiologists more accurately classify between benign and malignant lesions. The purpose of this study is to investigate a new mammographic mass classification scheme based on a deep learning method. In this study, we used an image dataset involving 560 regions of interest (ROIs) extracted from digital mammograms, which includes 280 malignant and 280 benign mass ROIs, respectively. An eight layer deep learning network was applied, which employs three pairs of convolution-max-pooling layers for automatic feature extraction and a multiple layer perception (MLP) classifier for feature categorization. In order to improve robustness of selected features, each convolution layer is connected with a max-pooling layer. A number of 20, 10, and 5 feature maps were utilized for the 1st, 2nd and 3rd convolution layer, respectively. The convolution networks are followed by a MLP classifier, which generates a classification score to predict likelihood of a ROI depicting a malignant mass. Among 560 ROIs, 420 ROIs were used as a training dataset and the remaining 140 ROIs were used as a validation dataset. The result shows that the new deep learning based classifier yielded an area under the receiver operation characteristic curve (AUC) of 0.810+/-0.036. This study demonstrated the potential superiority of using a deep learning based classifier to distinguish malignant and benign breast masses without segmenting the lesions and extracting the pre-defined image features.

Improving our understanding of hydraulic-electrical relations: A case study of the surficial aquifer in Emirate Abu Dhabi

USGS Publications Warehouse

Ikard, Scott; Kress, Wade

2016-01-01

Transmissivity is a bulk hydraulic property that can be correlated with bulk electrical properties of an aquifer. In aquifers that are electrically-resistive relative to adjacent layers in a horizontally stratified sequence, transmissivity has been shown to correlate with bulk transverse resistance. Conversely, in aquifers that are electrically-conductive relative to adjacent layers, transmissivity has been shown to correlate with bulk longitudinal conductance. In both cases, previous investigations have relied on small datasets (on average less than eight observations) that have yielded coefficients of determination (R2) that are typically in the range of 0.6 to 0.7 to substantiate these relations. Compared to previous investigations, this paper explores hydraulic-electrical relations using a much larger dataset. Geophysical data collected from 26 boreholes in Emirate Abu Dhabi, United Arab Emirates, are used to correlate transmissivity modeled from neutron porosity logs to the bulk electrical properties of the surficial aquifer that are computed from deep-induction logs. Transmissivity is found to be highly correlated with longitudinal conductance. An R2 value of 0.853 is obtained when electrical effects caused by variations in pore-fluid salinity are taken into consideration.

DNApod: DNA polymorphism annotation database from next-generation sequence read archives.

PubMed

Mochizuki, Takako; Tanizawa, Yasuhiro; Fujisawa, Takatomo; Ohta, Tazro; Nikoh, Naruo; Shimizu, Tokurou; Toyoda, Atsushi; Fujiyama, Asao; Kurata, Nori; Nagasaki, Hideki; Kaminuma, Eli; Nakamura, Yasukazu

2017-01-01

With the rapid advances in next-generation sequencing (NGS), datasets for DNA polymorphisms among various species and strains have been produced, stored, and distributed. However, reliability varies among these datasets because the experimental and analytical conditions used differ among assays. Furthermore, such datasets have been frequently distributed from the websites of individual sequencing projects. It is desirable to integrate DNA polymorphism data into one database featuring uniform quality control that is distributed from a single platform at a single place. DNA polymorphism annotation database (DNApod; http://tga.nig.ac.jp/dnapod/) is an integrated database that stores genome-wide DNA polymorphism datasets acquired under uniform analytical conditions, and this includes uniformity in the quality of the raw data, the reference genome version, and evaluation algorithms. DNApod genotypic data are re-analyzed whole-genome shotgun datasets extracted from sequence read archives, and DNApod distributes genome-wide DNA polymorphism datasets and known-gene annotations for each DNA polymorphism. This new database was developed for storing genome-wide DNA polymorphism datasets of plants, with crops being the first priority. Here, we describe our analyzed data for 679, 404, and 66 strains of rice, maize, and sorghum, respectively. The analytical methods are available as a DNApod workflow in an NGS annotation system of the DNA Data Bank of Japan and a virtual machine image. Furthermore, DNApod provides tables of links of identifiers between DNApod genotypic data and public phenotypic data. To advance the sharing of organism knowledge, DNApod offers basic and ubiquitous functions for multiple alignment and phylogenetic tree construction by using orthologous gene information.

DNApod: DNA polymorphism annotation database from next-generation sequence read archives

PubMed Central

Mochizuki, Takako; Tanizawa, Yasuhiro; Fujisawa, Takatomo; Ohta, Tazro; Nikoh, Naruo; Shimizu, Tokurou; Toyoda, Atsushi; Fujiyama, Asao; Kurata, Nori; Nagasaki, Hideki; Kaminuma, Eli; Nakamura, Yasukazu

2017-01-01

With the rapid advances in next-generation sequencing (NGS), datasets for DNA polymorphisms among various species and strains have been produced, stored, and distributed. However, reliability varies among these datasets because the experimental and analytical conditions used differ among assays. Furthermore, such datasets have been frequently distributed from the websites of individual sequencing projects. It is desirable to integrate DNA polymorphism data into one database featuring uniform quality control that is distributed from a single platform at a single place. DNA polymorphism annotation database (DNApod; http://tga.nig.ac.jp/dnapod/) is an integrated database that stores genome-wide DNA polymorphism datasets acquired under uniform analytical conditions, and this includes uniformity in the quality of the raw data, the reference genome version, and evaluation algorithms. DNApod genotypic data are re-analyzed whole-genome shotgun datasets extracted from sequence read archives, and DNApod distributes genome-wide DNA polymorphism datasets and known-gene annotations for each DNA polymorphism. This new database was developed for storing genome-wide DNA polymorphism datasets of plants, with crops being the first priority. Here, we describe our analyzed data for 679, 404, and 66 strains of rice, maize, and sorghum, respectively. The analytical methods are available as a DNApod workflow in an NGS annotation system of the DNA Data Bank of Japan and a virtual machine image. Furthermore, DNApod provides tables of links of identifiers between DNApod genotypic data and public phenotypic data. To advance the sharing of organism knowledge, DNApod offers basic and ubiquitous functions for multiple alignment and phylogenetic tree construction by using orthologous gene information. PMID:28234924

Toward a real-time system for temporal enhanced ultrasound-guided prostate biopsy.

PubMed

Azizi, Shekoofeh; Van Woudenberg, Nathan; Sojoudi, Samira; Li, Ming; Xu, Sheng; Abu Anas, Emran M; Yan, Pingkun; Tahmasebi, Amir; Kwak, Jin Tae; Turkbey, Baris; Choyke, Peter; Pinto, Peter; Wood, Bradford; Mousavi, Parvin; Abolmaesumi, Purang

2018-03-27

We have previously proposed temporal enhanced ultrasound (TeUS) as a new paradigm for tissue characterization. TeUS is based on analyzing a sequence of ultrasound data with deep learning and has been demonstrated to be successful for detection of cancer in ultrasound-guided prostate biopsy. Our aim is to enable the dissemination of this technology to the community for large-scale clinical validation. In this paper, we present a unified software framework demonstrating near-real-time analysis of ultrasound data stream using a deep learning solution. The system integrates ultrasound imaging hardware, visualization and a deep learning back-end to build an accessible, flexible and robust platform. A client-server approach is used in order to run computationally expensive algorithms in parallel. We demonstrate the efficacy of the framework using two applications as case studies. First, we show that prostate cancer detection using near-real-time analysis of RF and B-mode TeUS data and deep learning is feasible. Second, we present real-time segmentation of ultrasound prostate data using an integrated deep learning solution. The system is evaluated for cancer detection accuracy on ultrasound data obtained from a large clinical study with 255 biopsy cores from 157 subjects. It is further assessed with an independent dataset with 21 biopsy targets from six subjects. In the first study, we achieve area under the curve, sensitivity, specificity and accuracy of 0.94, 0.77, 0.94 and 0.92, respectively, for the detection of prostate cancer. In the second study, we achieve an AUC of 0.85. Our results suggest that TeUS-guided biopsy can be potentially effective for the detection of prostate cancer.

TaxI: a software tool for DNA barcoding using distance methods

PubMed Central

Steinke, Dirk; Vences, Miguel; Salzburger, Walter; Meyer, Axel

2005-01-01

DNA barcoding is a promising approach to the diagnosis of biological diversity in which DNA sequences serve as the primary key for information retrieval. Most existing software for evolutionary analysis of DNA sequences was designed for phylogenetic analyses and, hence, those algorithms do not offer appropriate solutions for the rapid, but precise analyses needed for DNA barcoding, and are also unable to process the often large comparative datasets. We developed a flexible software tool for DNA taxonomy, named TaxI. This program calculates sequence divergences between a query sequence (taxon to be barcoded) and each sequence of a dataset of reference sequences defined by the user. Because the analysis is based on separate pairwise alignments this software is also able to work with sequences characterized by multiple insertions and deletions that are difficult to align in large sequence sets (i.e. thousands of sequences) by multiple alignment algorithms because of computational restrictions. Here, we demonstrate the utility of this approach with two datasets of fish larvae and juveniles from Lake Constance and juvenile land snails under different models of sequence evolution. Sets of ribosomal 16S rRNA sequences, characterized by multiple indels, performed as good as or better than cox1 sequence sets in assigning sequences to species, demonstrating the suitability of rRNA genes for DNA barcoding. PMID:16214755

DNA demethylation activates genes in seed maternal integument development in rice (Oryza sativa L.).

PubMed

Wang, Yifeng; Lin, Haiyan; Tong, Xiaohong; Hou, Yuxuan; Chang, Yuxiao; Zhang, Jian

2017-11-01

DNA methylation is an important epigenetic modification that regulates various plant developmental processes. Rice seed integument determines the seed size. However, the role of DNA methylation in its development remains largely unknown. Here, we report the first dynamic DNA methylomic profiling of rice maternal integument before and after pollination by using a whole-genome bisulfite deep sequencing approach. Analysis of DNA methylation patterns identified 4238 differentially methylated regions underpin 4112 differentially methylated genes, including GW2, DEP1, RGB1 and numerous other regulators participated in maternal integument development. Bisulfite sanger sequencing and qRT-PCR of six differentially methylated genes revealed extensive occurrence of DNA hypomethylation triggered by double fertilization at IAP compared with IBP, suggesting that DNA demethylation might be a key mechanism to activate numerous maternal controlling genes. These results presented here not only greatly expanded the rice methylome dataset, but also shed novel insight into the regulatory roles of DNA methylation in rice seed maternal integument development. Copyright © 2017 Elsevier Masson SAS. All rights reserved.

Processing and population genetic analysis of multigenic datasets with ProSeq3 software.

PubMed

Filatov, Dmitry A

2009-12-01

The current tendency in molecular population genetics is to use increasing numbers of genes in the analysis. Here I describe a program for handling and population genetic analysis of DNA polymorphism data collected from multiple genes. The program includes a sequence/alignment editor and an internal relational database that simplify the preparation and manipulation of multigenic DNA polymorphism datasets. The most commonly used DNA polymorphism analyses are implemented in ProSeq3, facilitating population genetic analysis of large multigenic datasets. Extensive input/output options make ProSeq3 a convenient hub for sequence data processing and analysis. The program is available free of charge from http://dps.plants.ox.ac.uk/sequencing/proseq.htm.

The MIND PALACE: A Multi-Spectral Imaging and Spectroscopy Database for Planetary Science

NASA Astrophysics Data System (ADS)

Eshelman, E.; Doloboff, I.; Hara, E. K.; Uckert, K.; Sapers, H. M.; Abbey, W.; Beegle, L. W.; Bhartia, R.

2017-12-01

The Multi-Instrument Database (MIND) is the web-based home to a well-characterized set of analytical data collected by a suite of deep-UV fluorescence/Raman instruments built at the Jet Propulsion Laboratory (JPL). Samples derive from a growing body of planetary surface analogs, mineral and microbial standards, meteorites, spacecraft materials, and other astrobiologically relevant materials. In addition to deep-UV spectroscopy, datasets stored in MIND are obtained from a variety of analytical techniques obtained over multiple spatial and spectral scales including electron microscopy, optical microscopy, infrared spectroscopy, X-ray fluorescence, and direct fluorescence imaging. Multivariate statistical analysis techniques, primarily Principal Component Analysis (PCA), are used to guide interpretation of these large multi-analytical spectral datasets. Spatial co-referencing of integrated spectral/visual maps is performed using QGIS (geographic information system software). Georeferencing techniques transform individual instrument data maps into a layered co-registered data cube for analysis across spectral and spatial scales. The body of data in MIND is intended to serve as a permanent, reliable, and expanding database of deep-UV spectroscopy datasets generated by this unique suite of JPL-based instruments on samples of broad planetary science interest.

Identification of autism spectrum disorder using deep learning and the ABIDE dataset.

PubMed

Heinsfeld, Anibal Sólon; Franco, Alexandre Rosa; Craddock, R Cameron; Buchweitz, Augusto; Meneguzzi, Felipe

2018-01-01

The goal of the present study was to apply deep learning algorithms to identify autism spectrum disorder (ASD) patients from large brain imaging dataset, based solely on the patients brain activation patterns. We investigated ASD patients brain imaging data from a world-wide multi-site database known as ABIDE (Autism Brain Imaging Data Exchange). ASD is a brain-based disorder characterized by social deficits and repetitive behaviors. According to recent Centers for Disease Control data, ASD affects one in 68 children in the United States. We investigated patterns of functional connectivity that objectively identify ASD participants from functional brain imaging data, and attempted to unveil the neural patterns that emerged from the classification. The results improved the state-of-the-art by achieving 70% accuracy in identification of ASD versus control patients in the dataset. The patterns that emerged from the classification show an anticorrelation of brain function between anterior and posterior areas of the brain; the anticorrelation corroborates current empirical evidence of anterior-posterior disruption in brain connectivity in ASD. We present the results and identify the areas of the brain that contributed most to differentiating ASD from typically developing controls as per our deep learning model.

Cascaded deep decision networks for classification of endoscopic images

NASA Astrophysics Data System (ADS)

Murthy, Venkatesh N.; Singh, Vivek; Sun, Shanhui; Bhattacharya, Subhabrata; Chen, Terrence; Comaniciu, Dorin

2017-02-01

Both traditional and wireless capsule endoscopes can generate tens of thousands of images for each patient. It is desirable to have the majority of irrelevant images filtered out by automatic algorithms during an offline review process or to have automatic indication for highly suspicious areas during an online guidance. This also applies to the newly invented endomicroscopy, where online indication of tumor classification plays a significant role. Image classification is a standard pattern recognition problem and is well studied in the literature. However, performance on the challenging endoscopic images still has room for improvement. In this paper, we present a novel Cascaded Deep Decision Network (CDDN) to improve image classification performance over standard Deep neural network based methods. During the learning phase, CDDN automatically builds a network which discards samples that are classified with high confidence scores by a previously trained network and concentrates only on the challenging samples which would be handled by the subsequent expert shallow networks. We validate CDDN using two different types of endoscopic imaging, which includes a polyp classification dataset and a tumor classification dataset. From both datasets we show that CDDN can outperform other methods by about 10%. In addition, CDDN can also be applied to other image classification problems.

Sleep spindle detection using deep learning: A validation study based on crowdsourcing.

PubMed

Dakun Tan; Rui Zhao; Jinbo Sun; Wei Qin

2015-08-01

Sleep spindles are significant transient oscillations observed on the electroencephalogram (EEG) in stage 2 of non-rapid eye movement sleep. Deep belief network (DBN) gaining great successes in images and speech is still a novel method to develop sleep spindle detection system. In this paper, crowdsourcing replacing gold standard was applied to generate three different labeled samples and constructed three classes of datasets with a combination of these samples. An F1-score measure was estimated to compare the performance of DBN to other three classifiers on classifying these samples, with the DBN obtaining an result of 92.78%. Then a comparison of two feature extraction methods based on power spectrum density was made on same dataset using DBN. In addition, the DBN trained in dataset was applied to detect sleep spindle from raw EEG recordings and performed a comparable capacity to expert group consensus.

GeneImp: Fast Imputation to Large Reference Panels Using Genotype Likelihoods from Ultralow Coverage Sequencing

PubMed Central

Spiliopoulou, Athina; Colombo, Marco; Orchard, Peter; Agakov, Felix; McKeigue, Paul

2017-01-01

We address the task of genotype imputation to a dense reference panel given genotype likelihoods computed from ultralow coverage sequencing as inputs. In this setting, the data have a high-level of missingness or uncertainty, and are thus more amenable to a probabilistic representation. Most existing imputation algorithms are not well suited for this situation, as they rely on prephasing for computational efficiency, and, without definite genotype calls, the prephasing task becomes computationally expensive. We describe GeneImp, a program for genotype imputation that does not require prephasing and is computationally tractable for whole-genome imputation. GeneImp does not explicitly model recombination, instead it capitalizes on the existence of large reference panels—comprising thousands of reference haplotypes—and assumes that the reference haplotypes can adequately represent the target haplotypes over short regions unaltered. We validate GeneImp based on data from ultralow coverage sequencing (0.5×), and compare its performance to the most recent version of BEAGLE that can perform this task. We show that GeneImp achieves imputation quality very close to that of BEAGLE, using one to two orders of magnitude less time, without an increase in memory complexity. Therefore, GeneImp is the first practical choice for whole-genome imputation to a dense reference panel when prephasing cannot be applied, for instance, in datasets produced via ultralow coverage sequencing. A related future application for GeneImp is whole-genome imputation based on the off-target reads from deep whole-exome sequencing. PMID:28348060

Comparative Evaluation of Background Subtraction Algorithms in Remote Scene Videos Captured by MWIR Sensors

PubMed Central

Yao, Guangle; Lei, Tao; Zhong, Jiandan; Jiang, Ping; Jia, Wenwu

2017-01-01

Background subtraction (BS) is one of the most commonly encountered tasks in video analysis and tracking systems. It distinguishes the foreground (moving objects) from the video sequences captured by static imaging sensors. Background subtraction in remote scene infrared (IR) video is important and common to lots of fields. This paper provides a Remote Scene IR Dataset captured by our designed medium-wave infrared (MWIR) sensor. Each video sequence in this dataset is identified with specific BS challenges and the pixel-wise ground truth of foreground (FG) for each frame is also provided. A series of experiments were conducted to evaluate BS algorithms on this proposed dataset. The overall performance of BS algorithms and the processor/memory requirements were compared. Proper evaluation metrics or criteria were employed to evaluate the capability of each BS algorithm to handle different kinds of BS challenges represented in this dataset. The results and conclusions in this paper provide valid references to develop new BS algorithm for remote scene IR video sequence, and some of them are not only limited to remote scene or IR video sequence but also generic for background subtraction. The Remote Scene IR dataset and the foreground masks detected by each evaluated BS algorithm are available online: https://github.com/JerryYaoGl/BSEvaluationRemoteSceneIR. PMID:28837112

Deep Sequencing of T-cell Receptor DNA as a Biomarker of Clonally Expanded TILs in Breast Cancer after Immunotherapy.

PubMed

Page, David B; Yuan, Jianda; Redmond, David; Wen, Y Hanna; Durack, Jeremy C; Emerson, Ryan; Solomon, Stephen; Dong, Zhiwan; Wong, Phillip; Comstock, Christopher; Diab, Adi; Sung, Janice; Maybody, Majid; Morris, Elizabeth; Brogi, Edi; Morrow, Monica; Sacchini, Virgilio; Elemento, Olivier; Robins, Harlan; Patil, Sujata; Allison, James P; Wolchok, Jedd D; Hudis, Clifford; Norton, Larry; McArthur, Heather L

2016-10-01

In early-stage breast cancer, the degree of tumor-infiltrating lymphocytes (TIL) predicts response to chemotherapy and overall survival. Combination immunotherapy with immune checkpoint antibody plus tumor cryoablation can induce lymphocytic infiltrates and improve survival in mice. We used T-cell receptor (TCR) DNA sequencing to evaluate both the effect of cryoimmunotherapy in humans and the feasibility of TCR sequencing in early-stage breast cancer. In a pilot clinical trial, 18 women with early-stage breast cancer were treated preoperatively with cryoablation, single-dose anti-CTLA-4 (ipilimumab), or cryoablation + ipilimumab. TCRs within serially collected peripheral blood and tumor tissue were sequenced. In baseline tumor tissues, T-cell density as measured by TCR sequencing correlated with TIL scores obtained by hematoxylin and eosin (H&E) staining. However, tumors with little or no lymphocytes by H&E contained up to 3.6 × 10 6 TCR DNA sequences, highlighting the sensitivity of the ImmunoSEQ platform. In this dataset, ipilimumab increased intratumoral T-cell density over time, whereas cryoablation ± ipilimumab diversified and remodeled the intratumoral T-cell clonal repertoire. Compared with monotherapy, cryoablation plus ipilimumab was associated with numerically greater numbers of peripheral blood and intratumoral T-cell clones expanding robustly following therapy. In conclusion, TCR sequencing correlates with H&E lymphocyte scoring and provides additional information on clonal diversity. These findings support further study of the use of TCR sequencing as a biomarker for T-cell responses to therapy and for the study of cryoimmunotherapy in early-stage breast cancer. Cancer Immunol Res; 4(10); 835-44. ©2016 AACR. ©2016 American Association for Cancer Research.

CPTAC Releases Largest-Ever Breast Cancer Proteome Dataset from Previously Genome Characterized Tumors | Office of Cancer Clinical Proteomics Research

Cancer.gov

National Cancer Institute (NCI) Clinical Proteomic Tumor Analysis Consortium (CPTAC) scientists have released a dataset of proteins and  phosphopeptides identified through deep proteomic and phosphoproteomic analysis of breast tumor samples, previously genomically analyzed by The Cancer Genome Atlas (TCGA).

CLUSTOM-CLOUD: In-Memory Data Grid-Based Software for Clustering 16S rRNA Sequence Data in the Cloud Environment.

PubMed

Oh, Jeongsu; Choi, Chi-Hwan; Park, Min-Kyu; Kim, Byung Kwon; Hwang, Kyuin; Lee, Sang-Heon; Hong, Soon Gyu; Nasir, Arshan; Cho, Wan-Sup; Kim, Kyung Mo

2016-01-01

High-throughput sequencing can produce hundreds of thousands of 16S rRNA sequence reads corresponding to different organisms present in the environmental samples. Typically, analysis of microbial diversity in bioinformatics starts from pre-processing followed by clustering 16S rRNA reads into relatively fewer operational taxonomic units (OTUs). The OTUs are reliable indicators of microbial diversity and greatly accelerate the downstream analysis time. However, existing hierarchical clustering algorithms that are generally more accurate than greedy heuristic algorithms struggle with large sequence datasets. To keep pace with the rapid rise in sequencing data, we present CLUSTOM-CLOUD, which is the first distributed sequence clustering program based on In-Memory Data Grid (IMDG) technology-a distributed data structure to store all data in the main memory of multiple computing nodes. The IMDG technology helps CLUSTOM-CLOUD to enhance both its capability of handling larger datasets and its computational scalability better than its ancestor, CLUSTOM, while maintaining high accuracy. Clustering speed of CLUSTOM-CLOUD was evaluated on published 16S rRNA human microbiome sequence datasets using the small laboratory cluster (10 nodes) and under the Amazon EC2 cloud-computing environments. Under the laboratory environment, it required only ~3 hours to process dataset of size 200 K reads regardless of the complexity of the human microbiome data. In turn, one million reads were processed in approximately 20, 14, and 11 hours when utilizing 20, 30, and 40 nodes on the Amazon EC2 cloud-computing environment. The running time evaluation indicates that CLUSTOM-CLOUD can handle much larger sequence datasets than CLUSTOM and is also a scalable distributed processing system. The comparative accuracy test using 16S rRNA pyrosequences of a mock community shows that CLUSTOM-CLOUD achieves higher accuracy than DOTUR, mothur, ESPRIT-Tree, UCLUST and Swarm. CLUSTOM-CLOUD is written in JAVA and is freely available at http://clustomcloud.kopri.re.kr.

CLUSTOM-CLOUD: In-Memory Data Grid-Based Software for Clustering 16S rRNA Sequence Data in the Cloud Environment

PubMed Central

Park, Min-Kyu; Kim, Byung Kwon; Hwang, Kyuin; Lee, Sang-Heon; Hong, Soon Gyu; Nasir, Arshan; Cho, Wan-Sup; Kim, Kyung Mo

2016-01-01

High-throughput sequencing can produce hundreds of thousands of 16S rRNA sequence reads corresponding to different organisms present in the environmental samples. Typically, analysis of microbial diversity in bioinformatics starts from pre-processing followed by clustering 16S rRNA reads into relatively fewer operational taxonomic units (OTUs). The OTUs are reliable indicators of microbial diversity and greatly accelerate the downstream analysis time. However, existing hierarchical clustering algorithms that are generally more accurate than greedy heuristic algorithms struggle with large sequence datasets. To keep pace with the rapid rise in sequencing data, we present CLUSTOM-CLOUD, which is the first distributed sequence clustering program based on In-Memory Data Grid (IMDG) technology–a distributed data structure to store all data in the main memory of multiple computing nodes. The IMDG technology helps CLUSTOM-CLOUD to enhance both its capability of handling larger datasets and its computational scalability better than its ancestor, CLUSTOM, while maintaining high accuracy. Clustering speed of CLUSTOM-CLOUD was evaluated on published 16S rRNA human microbiome sequence datasets using the small laboratory cluster (10 nodes) and under the Amazon EC2 cloud-computing environments. Under the laboratory environment, it required only ~3 hours to process dataset of size 200 K reads regardless of the complexity of the human microbiome data. In turn, one million reads were processed in approximately 20, 14, and 11 hours when utilizing 20, 30, and 40 nodes on the Amazon EC2 cloud-computing environment. The running time evaluation indicates that CLUSTOM-CLOUD can handle much larger sequence datasets than CLUSTOM and is also a scalable distributed processing system. The comparative accuracy test using 16S rRNA pyrosequences of a mock community shows that CLUSTOM-CLOUD achieves higher accuracy than DOTUR, mothur, ESPRIT-Tree, UCLUST and Swarm. CLUSTOM-CLOUD is written in JAVA and is freely available at http://clustomcloud.kopri.re.kr. PMID:26954507

Low-Latency Telerobotic Sample Return and Biomolecular Sequencing for Deep Space Gateway

NASA Astrophysics Data System (ADS)

Lupisella, M.; Bleacher, J.; Lewis, R.; Dworkin, J.; Wright, M.; Burton, A.; Rubins, K.; Wallace, S.; Stahl, S.; John, K.; Archer, D.; Niles, P.; Regberg, A.; Smith, D.; Race, M.; Chiu, C.; Russell, J.; Rampe, E.; Bywaters, K.

2018-02-01

Low-latency telerobotics, crew-assisted sample return, and biomolecular sequencing can be used to acquire and analyze lunar farside and/or Apollo landing site samples. Sequencing can also be used to monitor and study Deep Space Gateway environment and crew health.

Deep Convolutional Neural Networks for Computer-Aided Detection: CNN Architectures, Dataset Characteristics and Transfer Learning.

PubMed

Shin, Hoo-Chang; Roth, Holger R; Gao, Mingchen; Lu, Le; Xu, Ziyue; Nogues, Isabella; Yao, Jianhua; Mollura, Daniel; Summers, Ronald M

2016-05-01

Remarkable progress has been made in image recognition, primarily due to the availability of large-scale annotated datasets and deep convolutional neural networks (CNNs). CNNs enable learning data-driven, highly representative, hierarchical image features from sufficient training data. However, obtaining datasets as comprehensively annotated as ImageNet in the medical imaging domain remains a challenge. There are currently three major techniques that successfully employ CNNs to medical image classification: training the CNN from scratch, using off-the-shelf pre-trained CNN features, and conducting unsupervised CNN pre-training with supervised fine-tuning. Another effective method is transfer learning, i.e., fine-tuning CNN models pre-trained from natural image dataset to medical image tasks. In this paper, we exploit three important, but previously understudied factors of employing deep convolutional neural networks to computer-aided detection problems. We first explore and evaluate different CNN architectures. The studied models contain 5 thousand to 160 million parameters, and vary in numbers of layers. We then evaluate the influence of dataset scale and spatial image context on performance. Finally, we examine when and why transfer learning from pre-trained ImageNet (via fine-tuning) can be useful. We study two specific computer-aided detection (CADe) problems, namely thoraco-abdominal lymph node (LN) detection and interstitial lung disease (ILD) classification. We achieve the state-of-the-art performance on the mediastinal LN detection, and report the first five-fold cross-validation classification results on predicting axial CT slices with ILD categories. Our extensive empirical evaluation, CNN model analysis and valuable insights can be extended to the design of high performance CAD systems for other medical imaging tasks.

Validation and Uncertainty Estimates for MODIS Collection 6 "Deep Blue" Aerosol Data

NASA Technical Reports Server (NTRS)

Sayer, A. M.; Hsu, N. C.; Bettenhausen, C.; Jeong, M.-J.

2013-01-01

The "Deep Blue" aerosol optical depth (AOD) retrieval algorithm was introduced in Collection 5 of the Moderate Resolution Imaging Spectroradiometer (MODIS) product suite, and complemented the existing "Dark Target" land and ocean algorithms by retrieving AOD over bright arid land surfaces, such as deserts. The forthcoming Collection 6 of MODIS products will include a "second generation" Deep Blue algorithm, expanding coverage to all cloud-free and snow-free land surfaces. The Deep Blue dataset will also provide an estimate of the absolute uncertainty on AOD at 550 nm for each retrieval. This study describes the validation of Deep Blue Collection 6 AOD at 550 nm (Tau(sub M)) from MODIS Aqua against Aerosol Robotic Network (AERONET) data from 60 sites to quantify these uncertainties. The highest quality (denoted quality assurance flag value 3) data are shown to have an absolute uncertainty of approximately (0.086+0.56Tau(sub M))/AMF, where AMF is the geometric air mass factor. For a typical AMF of 2.8, this is approximately 0.03+0.20Tau(sub M), comparable in quality to other satellite AOD datasets. Regional variability of retrieval performance and comparisons against Collection 5 results are also discussed.

DNCON2: improved protein contact prediction using two-level deep convolutional neural networks.

PubMed

Adhikari, Badri; Hou, Jie; Cheng, Jianlin

2018-05-01

Significant improvements in the prediction of protein residue-residue contacts are observed in the recent years. These contacts, predicted using a variety of coevolution-based and machine learning methods, are the key contributors to the recent progress in ab initio protein structure prediction, as demonstrated in the recent CASP experiments. Continuing the development of new methods to reliably predict contact maps is essential to further improve ab initio structure prediction. In this paper we discuss DNCON2, an improved protein contact map predictor based on two-level deep convolutional neural networks. It consists of six convolutional neural networks-the first five predict contacts at 6, 7.5, 8, 8.5 and 10 Å distance thresholds, and the last one uses these five predictions as additional features to predict final contact maps. On the free-modeling datasets in CASP10, 11 and 12 experiments, DNCON2 achieves mean precisions of 35, 50 and 53.4%, respectively, higher than 30.6% by MetaPSICOV on CASP10 dataset, 34% by MetaPSICOV on CASP11 dataset and 46.3% by Raptor-X on CASP12 dataset, when top L/5 long-range contacts are evaluated. We attribute the improved performance of DNCON2 to the inclusion of short- and medium-range contacts into training, two-level approach to prediction, use of the state-of-the-art optimization and activation functions, and a novel deep learning architecture that allows each filter in a convolutional layer to access all the input features of a protein of arbitrary length. The web server of DNCON2 is at http://sysbio.rnet.missouri.edu/dncon2/ where training and testing datasets as well as the predictions for CASP10, 11 and 12 free-modeling datasets can also be downloaded. Its source code is available at https://github.com/multicom-toolbox/DNCON2/. chengji@missouri.edu. Supplementary data are available at Bioinformatics online.

Deep Convolutional Neural Networks for Computer-Aided Detection: CNN Architectures, Dataset Characteristics and Transfer Learning

PubMed Central

Hoo-Chang, Shin; Roth, Holger R.; Gao, Mingchen; Lu, Le; Xu, Ziyue; Nogues, Isabella; Yao, Jianhua; Mollura, Daniel

2016-01-01

Remarkable progress has been made in image recognition, primarily due to the availability of large-scale annotated datasets (i.e. ImageNet) and the revival of deep convolutional neural networks (CNN). CNNs enable learning data-driven, highly representative, layered hierarchical image features from sufficient training data. However, obtaining datasets as comprehensively annotated as ImageNet in the medical imaging domain remains a challenge. There are currently three major techniques that successfully employ CNNs to medical image classification: training the CNN from scratch, using off-the-shelf pre-trained CNN features, and conducting unsupervised CNN pre-training with supervised fine-tuning. Another effective method is transfer learning, i.e., fine-tuning CNN models (supervised) pre-trained from natural image dataset to medical image tasks (although domain transfer between two medical image datasets is also possible). In this paper, we exploit three important, but previously understudied factors of employing deep convolutional neural networks to computer-aided detection problems. We first explore and evaluate different CNN architectures. The studied models contain 5 thousand to 160 million parameters, and vary in numbers of layers. We then evaluate the influence of dataset scale and spatial image context on performance. Finally, we examine when and why transfer learning from pre-trained ImageNet (via fine-tuning) can be useful. We study two specific computeraided detection (CADe) problems, namely thoraco-abdominal lymph node (LN) detection and interstitial lung disease (ILD) classification. We achieve the state-of-the-art performance on the mediastinal LN detection, with 85% sensitivity at 3 false positive per patient, and report the first five-fold cross-validation classification results on predicting axial CT slices with ILD categories. Our extensive empirical evaluation, CNN model analysis and valuable insights can be extended to the design of high performance CAD systems for other medical imaging tasks. PMID:26886976

Sequencing Data Discovery and Integration for Earth System Science with MetaSeek

NASA Astrophysics Data System (ADS)

Hoarfrost, A.; Brown, N.; Arnosti, C.

2017-12-01

Microbial communities play a central role in biogeochemical cycles. Sequencing data resources from environmental sources have grown exponentially in recent years, and represent a singular opportunity to investigate microbial interactions with Earth system processes. Carrying out such meta-analyses depends on our ability to discover and curate sequencing data into large-scale integrated datasets. However, such integration efforts are currently challenging and time-consuming, with sequencing data scattered across multiple repositories and metadata that is not easily or comprehensively searchable. MetaSeek is a sequencing data discovery tool that integrates sequencing metadata from all the major data repositories, allowing the user to search and filter on datasets in a lightweight application with an intuitive, easy-to-use web-based interface. Users can save and share curated datasets, while other users can browse these data integrations or use them as a jumping off point for their own curation. Missing and/or erroneous metadata are inferred automatically where possible, and where not possible, users are prompted to contribute to the improvement of the sequencing metadata pool by correcting and amending metadata errors. Once an integrated dataset has been curated, users can follow simple instructions to download their raw data and quickly begin their investigations. In addition to the online interface, the MetaSeek database is easily queryable via an open API, further enabling users and facilitating integrations of MetaSeek with other data curation tools. This tool lowers the barriers to curation and integration of environmental sequencing data, clearing the path forward to illuminating the ecosystem-scale interactions between biological and abiotic processes.

Fungal communities from the calcareous deep-sea sediments in the Southwest India Ridge revealed by Illumina sequencing technology.

PubMed

Zhang, Likui; Kang, Manyu; Huang, Yangchao; Yang, Lixiang

2016-05-01

The diversity and ecological significance of bacteria and archaea in deep-sea environments have been thoroughly investigated, but eukaryotic microorganisms in these areas, such as fungi, are poorly understood. To elucidate fungal diversity in calcareous deep-sea sediments in the Southwest India Ridge (SWIR), the internal transcribed spacer (ITS) regions of rRNA genes from two sediment metagenomic DNA samples were amplified and sequenced using the Illumina sequencing platform. The results revealed that 58-63 % and 36-42 % of the ITS sequences (97 % similarity) belonged to Basidiomycota and Ascomycota, respectively. These findings suggest that Basidiomycota and Ascomycota are the predominant fungal phyla in the two samples. We also found that Agaricomycetes, Leotiomycetes, and Pezizomycetes were the major fungal classes in the two samples. At the species level, Thelephoraceae sp. and Phialocephala fortinii were major fungal species in the two samples. Despite the low relative abundance, unidentified fungal sequences were also observed in the two samples. Furthermore, we found that there were slight differences in fungal diversity between the two sediment samples, although both were collected from the SWIR. Thus, our results demonstrate that calcareous deep-sea sediments in the SWIR harbor diverse fungi, which augment the fungal groups in deep-sea sediments. This is the first report of fungal communities in calcareous deep-sea sediments in the SWIR revealed by Illumina sequencing.

Organic carbon accumulation in modern sediments of the Angola basin influenced by the Congo deep-sea fan

NASA Astrophysics Data System (ADS)

Baudin, François; Martinez, Philippe; Dennielou, Bernard; Charlier, Karine; Marsset, Tania; Droz, Laurence; Rabouille, Christophe

2017-08-01

Geochemical data (total organic carbon-TOC content, δ13Corg, C:N, Rock-Eval analyses) were obtained on 150 core tops from the Angola basin, with a special focus on the Congo deep-sea fan. Combined with the previously published data, the resulting dataset (322 stations) shows a good spatial and bathymetric representativeness. TOC content and δ13Corg maps of the Angola basin were generated using this enhanced dataset. The main difference in our map with previously published ones is the high terrestrial organic matter content observed downslope along the active turbidite channel of the Congo deep-sea fan till the distal lobe complex near 5000 m of water-depth. Interpretation of downslope trends in TOC content and organic matter composition indicates that lateral particle transport by turbidity currents is the primary mechanism controlling supply and burial of organic matter in the bathypelagic depths.

MetaCAA: A clustering-aided methodology for efficient assembly of metagenomic datasets.

PubMed

Reddy, Rachamalla Maheedhar; Mohammed, Monzoorul Haque; Mande, Sharmila S

2014-01-01

A key challenge in analyzing metagenomics data pertains to assembly of sequenced DNA fragments (i.e. reads) originating from various microbes in a given environmental sample. Several existing methodologies can assemble reads originating from a single genome. However, these methodologies cannot be applied for efficient assembly of metagenomic sequence datasets. In this study, we present MetaCAA - a clustering-aided methodology which helps in improving the quality of metagenomic sequence assembly. MetaCAA initially groups sequences constituting a given metagenome into smaller clusters. Subsequently, sequences in each cluster are independently assembled using CAP3, an existing single genome assembly program. Contigs formed in each of the clusters along with the unassembled reads are then subjected to another round of assembly for generating the final set of contigs. Validation using simulated and real-world metagenomic datasets indicates that MetaCAA aids in improving the overall quality of assembly. A software implementation of MetaCAA is available at https://metagenomics.atc.tcs.com/MetaCAA. Copyright © 2014 Elsevier Inc. All rights reserved.

Position-specific automated processing of V3 env ultra-deep pyrosequencing data for predicting HIV-1 tropism

PubMed Central

Jeanne, Nicolas; Saliou, Adrien; Carcenac, Romain; Lefebvre, Caroline; Dubois, Martine; Cazabat, Michelle; Nicot, Florence; Loiseau, Claire; Raymond, Stéphanie; Izopet, Jacques; Delobel, Pierre

2015-01-01

HIV-1 coreceptor usage must be accurately determined before starting CCR5 antagonist-based treatment as the presence of undetected minor CXCR4-using variants can cause subsequent virological failure. Ultra-deep pyrosequencing of HIV-1 V3 env allows to detect low levels of CXCR4-using variants that current genotypic approaches miss. However, the computation of the mass of sequence data and the need to identify true minor variants while excluding artifactual sequences generated during amplification and ultra-deep pyrosequencing is rate-limiting. Arbitrary fixed cut-offs below which minor variants are discarded are currently used but the errors generated during ultra-deep pyrosequencing are sequence-dependant rather than random. We have developed an automated processing of HIV-1 V3 env ultra-deep pyrosequencing data that uses biological filters to discard artifactual or non-functional V3 sequences followed by statistical filters to determine position-specific sensitivity thresholds, rather than arbitrary fixed cut-offs. It allows to retain authentic sequences with point mutations at V3 positions of interest and discard artifactual ones with accurate sensitivity thresholds. PMID:26585833

Position-specific automated processing of V3 env ultra-deep pyrosequencing data for predicting HIV-1 tropism.

PubMed

Jeanne, Nicolas; Saliou, Adrien; Carcenac, Romain; Lefebvre, Caroline; Dubois, Martine; Cazabat, Michelle; Nicot, Florence; Loiseau, Claire; Raymond, Stéphanie; Izopet, Jacques; Delobel, Pierre

2015-11-20

HIV-1 coreceptor usage must be accurately determined before starting CCR5 antagonist-based treatment as the presence of undetected minor CXCR4-using variants can cause subsequent virological failure. Ultra-deep pyrosequencing of HIV-1 V3 env allows to detect low levels of CXCR4-using variants that current genotypic approaches miss. However, the computation of the mass of sequence data and the need to identify true minor variants while excluding artifactual sequences generated during amplification and ultra-deep pyrosequencing is rate-limiting. Arbitrary fixed cut-offs below which minor variants are discarded are currently used but the errors generated during ultra-deep pyrosequencing are sequence-dependant rather than random. We have developed an automated processing of HIV-1 V3 env ultra-deep pyrosequencing data that uses biological filters to discard artifactual or non-functional V3 sequences followed by statistical filters to determine position-specific sensitivity thresholds, rather than arbitrary fixed cut-offs. It allows to retain authentic sequences with point mutations at V3 positions of interest and discard artifactual ones with accurate sensitivity thresholds.

NutriNet: A Deep Learning Food and Drink Image Recognition System for Dietary Assessment.

PubMed

Mezgec, Simon; Koroušić Seljak, Barbara

2017-06-27

Automatic food image recognition systems are alleviating the process of food-intake estimation and dietary assessment. However, due to the nature of food images, their recognition is a particularly challenging task, which is why traditional approaches in the field have achieved a low classification accuracy. Deep neural networks have outperformed such solutions, and we present a novel approach to the problem of food and drink image detection and recognition that uses a newly-defined deep convolutional neural network architecture, called NutriNet. This architecture was tuned on a recognition dataset containing 225,953 512 × 512 pixel images of 520 different food and drink items from a broad spectrum of food groups, on which we achieved a classification accuracy of 86 . 72 % , along with an accuracy of 94 . 47 % on a detection dataset containing 130 , 517 images. We also performed a real-world test on a dataset of self-acquired images, combined with images from Parkinson's disease patients, all taken using a smartphone camera, achieving a top-five accuracy of 55 % , which is an encouraging result for real-world images. Additionally, we tested NutriNet on the University of Milano-Bicocca 2016 (UNIMIB2016) food image dataset, on which we improved upon the provided baseline recognition result. An online training component was implemented to continually fine-tune the food and drink recognition model on new images. The model is being used in practice as part of a mobile app for the dietary assessment of Parkinson's disease patients.

Application of a deep-learning method to the forecast of daily solar flare occurrence using Convolution Neural Network

NASA Astrophysics Data System (ADS)

Shin, Seulki; Moon, Yong-Jae; Chu, Hyoungseok

2017-08-01

As the application of deep-learning methods has been succeeded in various fields, they have a high potential to be applied to space weather forecasting. Convolutional neural network, one of deep learning methods, is specialized in image recognition. In this study, we apply the AlexNet architecture, which is a winner of Imagenet Large Scale Virtual Recognition Challenge (ILSVRC) 2012, to the forecast of daily solar flare occurrence using the MatConvNet software of MATLAB. Our input images are SOHO/MDI, EIT 195Å, and 304Å from January 1996 to December 2010, and output ones are yes or no of flare occurrence. We select training dataset from Jan 1996 to Dec 2000 and from Jan 2003 to Dec 2008. Testing dataset is chosen from Jan 2001 to Dec 2002 and from Jan 2009 to Dec 2010 in order to consider the solar cycle effect. In training dataset, we randomly select one fifth of training data for validation dataset to avoid the overfitting problem. Our model successfully forecasts the flare occurrence with about 0.90 probability of detection (POD) for common flares (C-, M-, and X-class). While POD of major flares (M- and X-class) forecasting is 0.96, false alarm rate (FAR) also scores relatively high(0.60). We also present several statistical parameters such as critical success index (CSI) and true skill statistics (TSS). Our model can immediately be applied to automatic forecasting service when image data are available.

Identification of fungi in shotgun metagenomics datasets

PubMed Central

Donovan, Paul D.; Gonzalez, Gabriel; Higgins, Desmond G.

2018-01-01

Metagenomics uses nucleic acid sequencing to characterize species diversity in different niches such as environmental biomes or the human microbiome. Most studies have used 16S rRNA amplicon sequencing to identify bacteria. However, the decreasing cost of sequencing has resulted in a gradual shift away from amplicon analyses and towards shotgun metagenomic sequencing. Shotgun metagenomic data can be used to identify a wide range of species, but have rarely been applied to fungal identification. Here, we develop a sequence classification pipeline, FindFungi, and use it to identify fungal sequences in public metagenome datasets. We focus primarily on animal metagenomes, especially those from pig and mouse microbiomes. We identified fungi in 39 of 70 datasets comprising 71 fungal species. At least 11 pathogenic species with zoonotic potential were identified, including Candida tropicalis. We identified Pseudogymnoascus species from 13 Antarctic soil samples initially analyzed for the presence of bacteria capable of degrading diesel oil. We also show that Candida tropicalis and Candida loboi are likely the same species. In addition, we identify several examples where contaminating DNA was erroneously included in fungal genome assemblies. PMID:29444186

Accurate identification of RNA editing sites from primitive sequence with deep neural networks.

PubMed

Ouyang, Zhangyi; Liu, Feng; Zhao, Chenghui; Ren, Chao; An, Gaole; Mei, Chuan; Bo, Xiaochen; Shu, Wenjie

2018-04-16

RNA editing is a post-transcriptional RNA sequence alteration. Current methods have identified editing sites and facilitated research but require sufficient genomic annotations and prior-knowledge-based filtering steps, resulting in a cumbersome, time-consuming identification process. Moreover, these methods have limited generalizability and applicability in species with insufficient genomic annotations or in conditions of limited prior knowledge. We developed DeepRed, a deep learning-based method that identifies RNA editing from primitive RNA sequences without prior-knowledge-based filtering steps or genomic annotations. DeepRed achieved 98.1% and 97.9% area under the curve (AUC) in training and test sets, respectively. We further validated DeepRed using experimentally verified U87 cell RNA-seq data, achieving 97.9% positive predictive value (PPV). We demonstrated that DeepRed offers better prediction accuracy and computational efficiency than current methods with large-scale, mass RNA-seq data. We used DeepRed to assess the impact of multiple factors on editing identification with RNA-seq data from the Association of Biomolecular Resource Facilities and Sequencing Quality Control projects. We explored developmental RNA editing pattern changes during human early embryogenesis and evolutionary patterns in Drosophila species and the primate lineage using DeepRed. Our work illustrates DeepRed's state-of-the-art performance; it may decipher the hidden principles behind RNA editing, making editing detection convenient and effective.

A New Approach to Develop Computer-aided Diagnosis Scheme of Breast Mass Classification Using Deep Learning Technology

PubMed Central

Qiu, Yuchen; Yan, Shiju; Gundreddy, Rohith Reddy; Wang, Yunzhi; Cheng, Samuel; Liu, Hong; Zheng, Bin

2017-01-01

PURPOSE To develop and test a deep learning based computer-aided diagnosis (CAD) scheme of mammograms for classifying between malignant and benign masses. METHODS An image dataset involving 560 regions of interest (ROIs) extracted from digital mammograms was used. After down-sampling each ROI from 512×512 to 64×64 pixel size, we applied an 8 layer deep learning network that involves 3 pairs of convolution-max-pooling layers for automatic feature extraction and a multiple layer perceptron (MLP) classifier for feature categorization to process ROIs. The 3 pairs of convolution layers contain 20, 10, and 5 feature maps, respectively. Each convolution layer is connected with a max-pooling layer to improve the feature robustness. The output of the sixth layer is fully connected with a MLP classifier, which is composed of one hidden layer and one logistic regression layer. The network then generates a classification score to predict the likelihood of ROI depicting a malignant mass. A four-fold cross validation method was applied to train and test this deep learning network. RESULTS The results revealed that this CAD scheme yields an area under the receiver operation characteristic curve (AUC) of 0.696±0.044, 0.802±0.037, 0.836±0.036, and 0.822±0.035 for fold 1 to 4 testing datasets, respectively. The overall AUC of the entire dataset is 0.790±0.019. CONCLUSIONS This study demonstrates the feasibility of applying a deep learning based CAD scheme to classify between malignant and benign breast masses without a lesion segmentation, image feature computation and selection process. PMID:28436410

Multi-categorical deep learning neural network to classify retinal images: A pilot study employing small database.

PubMed

Choi, Joon Yul; Yoo, Tae Keun; Seo, Jeong Gi; Kwak, Jiyong; Um, Terry Taewoong; Rim, Tyler Hyungtaek

2017-01-01

Deep learning emerges as a powerful tool for analyzing medical images. Retinal disease detection by using computer-aided diagnosis from fundus image has emerged as a new method. We applied deep learning convolutional neural network by using MatConvNet for an automated detection of multiple retinal diseases with fundus photographs involved in STructured Analysis of the REtina (STARE) database. Dataset was built by expanding data on 10 categories, including normal retina and nine retinal diseases. The optimal outcomes were acquired by using a random forest transfer learning based on VGG-19 architecture. The classification results depended greatly on the number of categories. As the number of categories increased, the performance of deep learning models was diminished. When all 10 categories were included, we obtained results with an accuracy of 30.5%, relative classifier information (RCI) of 0.052, and Cohen's kappa of 0.224. Considering three integrated normal, background diabetic retinopathy, and dry age-related macular degeneration, the multi-categorical classifier showed accuracy of 72.8%, 0.283 RCI, and 0.577 kappa. In addition, several ensemble classifiers enhanced the multi-categorical classification performance. The transfer learning incorporated with ensemble classifier of clustering and voting approach presented the best performance with accuracy of 36.7%, 0.053 RCI, and 0.225 kappa in the 10 retinal diseases classification problem. First, due to the small size of datasets, the deep learning techniques in this study were ineffective to be applied in clinics where numerous patients suffering from various types of retinal disorders visit for diagnosis and treatment. Second, we found that the transfer learning incorporated with ensemble classifiers can improve the classification performance in order to detect multi-categorical retinal diseases. Further studies should confirm the effectiveness of algorithms with large datasets obtained from hospitals.

A new approach to develop computer-aided diagnosis scheme of breast mass classification using deep learning technology.

PubMed

Qiu, Yuchen; Yan, Shiju; Gundreddy, Rohith Reddy; Wang, Yunzhi; Cheng, Samuel; Liu, Hong; Zheng, Bin

2017-01-01

To develop and test a deep learning based computer-aided diagnosis (CAD) scheme of mammograms for classifying between malignant and benign masses. An image dataset involving 560 regions of interest (ROIs) extracted from digital mammograms was used. After down-sampling each ROI from 512×512 to 64×64 pixel size, we applied an 8 layer deep learning network that involves 3 pairs of convolution-max-pooling layers for automatic feature extraction and a multiple layer perceptron (MLP) classifier for feature categorization to process ROIs. The 3 pairs of convolution layers contain 20, 10, and 5 feature maps, respectively. Each convolution layer is connected with a max-pooling layer to improve the feature robustness. The output of the sixth layer is fully connected with a MLP classifier, which is composed of one hidden layer and one logistic regression layer. The network then generates a classification score to predict the likelihood of ROI depicting a malignant mass. A four-fold cross validation method was applied to train and test this deep learning network. The results revealed that this CAD scheme yields an area under the receiver operation characteristic curve (AUC) of 0.696±0.044, 0.802±0.037, 0.836±0.036, and 0.822±0.035 for fold 1 to 4 testing datasets, respectively. The overall AUC of the entire dataset is 0.790±0.019. This study demonstrates the feasibility of applying a deep learning based CAD scheme to classify between malignant and benign breast masses without a lesion segmentation, image feature computation and selection process.

Animal Viruses Probe dataset (AVPDS) for microarray-based diagnosis and identification of viruses.

PubMed

Yadav, Brijesh S; Pokhriyal, Mayank; Vasishtha, Dinesh P; Sharma, Bhaskar

2014-03-01

AVPDS (Animal Viruses Probe dataset) is a dataset of virus-specific and conserve oligonucleotides for identification and diagnosis of viruses infecting animals. The current dataset contain 20,619 virus specific probes for 833 viruses and their subtypes and 3,988 conserved probes for 146 viral genera. Dataset of virus specific probe has been divided into two fields namely virus name and probe sequence. Similarly conserved probes for virus genera table have genus, and subgroup within genus name and probe sequence. The subgroup within genus is artificially divided subgroups with no taxonomic significance and contains probes which identifies viruses in that specific subgroup of the genus. Using this dataset we have successfully diagnosed the first case of Newcastle disease virus in sheep and reported a mixed infection of Bovine viral diarrhea and Bovine herpesvirus in cattle. These dataset also contains probes which cross reacts across species experimentally though computationally they meet specifications. These probes have been marked. We hope that this dataset will be useful in microarray-based detection of viruses. The dataset can be accessed through the link https://dl.dropboxusercontent.com/u/94060831/avpds/HOME.html.

A deep learning method for lincRNA detection using auto-encoder algorithm.

PubMed

Yu, Ning; Yu, Zeng; Pan, Yi

2017-12-06

RNA sequencing technique (RNA-seq) enables scientists to develop novel data-driven methods for discovering more unidentified lincRNAs. Meantime, knowledge-based technologies are experiencing a potential revolution ignited by the new deep learning methods. By scanning the newly found data set from RNA-seq, scientists have found that: (1) the expression of lincRNAs appears to be regulated, that is, the relevance exists along the DNA sequences; (2) lincRNAs contain some conversed patterns/motifs tethered together by non-conserved regions. The two evidences give the reasoning for adopting knowledge-based deep learning methods in lincRNA detection. Similar to coding region transcription, non-coding regions are split at transcriptional sites. However, regulatory RNAs rather than message RNAs are generated. That is, the transcribed RNAs participate the biological process as regulatory units instead of generating proteins. Identifying these transcriptional regions from non-coding regions is the first step towards lincRNA recognition. The auto-encoder method achieves 100% and 92.4% prediction accuracy on transcription sites over the putative data sets. The experimental results also show the excellent performance of predictive deep neural network on the lincRNA data sets compared with support vector machine and traditional neural network. In addition, it is validated through the newly discovered lincRNA data set and one unreported transcription site is found by feeding the whole annotated sequences through the deep learning machine, which indicates that deep learning method has the extensive ability for lincRNA prediction. The transcriptional sequences of lincRNAs are collected from the annotated human DNA genome data. Subsequently, a two-layer deep neural network is developed for the lincRNA detection, which adopts the auto-encoder algorithm and utilizes different encoding schemes to obtain the best performance over intergenic DNA sequence data. Driven by those newly annotated lincRNA data, deep learning methods based on auto-encoder algorithm can exert their capability in knowledge learning in order to capture the useful features and the information correlation along DNA genome sequences for lincRNA detection. As our knowledge, this is the first application to adopt the deep learning techniques for identifying lincRNA transcription sequences.

IMG/M: integrated genome and metagenome comparative data analysis system

DOE PAGES

Chen, I-Min A.; Markowitz, Victor M.; Chu, Ken; ...

2016-10-13

The Integrated Microbial Genomes with Microbiome Samples (IMG/M: https://img.jgi.doe.gov/m/) system contains annotated DNA and RNA sequence data of (i) archaeal, bacterial, eukaryotic and viral genomes from cultured organisms, (ii) single cell genomes (SCG) and genomes from metagenomes (GFM) from uncultured archaea, bacteria and viruses and (iii) metagenomes from environmental, host associated and engineered microbiome samples. Sequence data are generated by DOE's Joint Genome Institute (JGI), submitted by individual scientists, or collected from public sequence data archives. Structural and functional annotation is carried out by JGI's genome and metagenome annotation pipelines. A variety of analytical and visualization tools provide support formore » examining and comparing IMG/M's datasets. IMG/M allows open access interactive analysis of publicly available datasets, while manual curation, submission and access to private datasets and computationally intensive workspace-based analysis require login/password access to its expert review(ER) companion system (IMG/M ER: https://img.jgi.doe.gov/ mer/). Since the last report published in the 2014 NAR Database Issue, IMG/M's dataset content has tripled in terms of number of datasets and overall protein coding genes, while its analysis tools have been extended to cope with the rapid growth in the number and size of datasets handled by the system.« less

IMG/M: integrated genome and metagenome comparative data analysis system

DOE Office of Scientific and Technical Information (OSTI.GOV)

Chen, I-Min A.; Markowitz, Victor M.; Chu, Ken

The Integrated Microbial Genomes with Microbiome Samples (IMG/M: https://img.jgi.doe.gov/m/) system contains annotated DNA and RNA sequence data of (i) archaeal, bacterial, eukaryotic and viral genomes from cultured organisms, (ii) single cell genomes (SCG) and genomes from metagenomes (GFM) from uncultured archaea, bacteria and viruses and (iii) metagenomes from environmental, host associated and engineered microbiome samples. Sequence data are generated by DOE's Joint Genome Institute (JGI), submitted by individual scientists, or collected from public sequence data archives. Structural and functional annotation is carried out by JGI's genome and metagenome annotation pipelines. A variety of analytical and visualization tools provide support formore » examining and comparing IMG/M's datasets. IMG/M allows open access interactive analysis of publicly available datasets, while manual curation, submission and access to private datasets and computationally intensive workspace-based analysis require login/password access to its expert review(ER) companion system (IMG/M ER: https://img.jgi.doe.gov/ mer/). Since the last report published in the 2014 NAR Database Issue, IMG/M's dataset content has tripled in terms of number of datasets and overall protein coding genes, while its analysis tools have been extended to cope with the rapid growth in the number and size of datasets handled by the system.« less

IMG/M: integrated genome and metagenome comparative data analysis system

PubMed Central

Chen, I-Min A.; Markowitz, Victor M.; Chu, Ken; Palaniappan, Krishna; Szeto, Ernest; Pillay, Manoj; Ratner, Anna; Huang, Jinghua; Andersen, Evan; Huntemann, Marcel; Varghese, Neha; Hadjithomas, Michalis; Tennessen, Kristin; Nielsen, Torben; Ivanova, Natalia N.; Kyrpides, Nikos C.

2017-01-01

The Integrated Microbial Genomes with Microbiome Samples (IMG/M: https://img.jgi.doe.gov/m/) system contains annotated DNA and RNA sequence data of (i) archaeal, bacterial, eukaryotic and viral genomes from cultured organisms, (ii) single cell genomes (SCG) and genomes from metagenomes (GFM) from uncultured archaea, bacteria and viruses and (iii) metagenomes from environmental, host associated and engineered microbiome samples. Sequence data are generated by DOE's Joint Genome Institute (JGI), submitted by individual scientists, or collected from public sequence data archives. Structural and functional annotation is carried out by JGI's genome and metagenome annotation pipelines. A variety of analytical and visualization tools provide support for examining and comparing IMG/M's datasets. IMG/M allows open access interactive analysis of publicly available datasets, while manual curation, submission and access to private datasets and computationally intensive workspace-based analysis require login/password access to its expert review (ER) companion system (IMG/M ER: https://img.jgi.doe.gov/mer/). Since the last report published in the 2014 NAR Database Issue, IMG/M's dataset content has tripled in terms of number of datasets and overall protein coding genes, while its analysis tools have been extended to cope with the rapid growth in the number and size of datasets handled by the system. PMID:27738135

Anomalous Diffusion Measured by a Twice-Refocused Spin Echo Pulse Sequence: Analysis Using Fractional Order Calculus

PubMed Central

2011-01-01

Purpose To theoretically develop and experimentally validate a formulism based on a fractional order calculus (FC) diffusion model to characterize anomalous diffusion in brain tissues measured with a twice-refocused spin-echo (TRSE) pulse sequence. Materials and Methods The FC diffusion model is the fractional order generalization of the Bloch-Torrey equation. Using this model, an analytical expression was derived to describe the diffusion-induced signal attenuation in a TRSE pulse sequence. To experimentally validate this expression, a set of diffusion-weighted (DW) images was acquired at 3 Tesla from healthy human brains using a TRSE sequence with twelve b-values ranging from 0 to 2,600 s/mm2. For comparison, DW images were also acquired using a Stejskal-Tanner diffusion gradient in a single-shot spin-echo echo planar sequence. For both datasets, a Levenberg-Marquardt fitting algorithm was used to extract three parameters: diffusion coefficient D, fractional order derivative in space β, and a spatial parameter μ (in units of μm). Using adjusted R-squared values and standard deviations, D, β and μ values and the goodness-of-fit in three specific regions of interest (ROI) in white matter, gray matter, and cerebrospinal fluid were evaluated for each of the two datasets. In addition, spatially resolved parametric maps were assessed qualitatively. Results The analytical expression for the TRSE sequence, derived from the FC diffusion model, accurately characterized the diffusion-induced signal loss in brain tissues at high b-values. In the selected ROIs, the goodness-of-fit and standard deviations for the TRSE dataset were comparable with the results obtained from the Stejskal-Tanner dataset, demonstrating the robustness of the FC model across multiple data acquisition strategies. Qualitatively, the D, β, and μ maps from the TRSE dataset exhibited fewer artifacts, reflecting the improved immunity to eddy currents. Conclusion The diffusion-induced signal attenuation in a TRSE pulse sequence can be described by an FC diffusion model at high b-values. This model performs equally well for data acquired from the human brain tissues with a TRSE pulse sequence or a conventional Stejskal-Tanner sequence. PMID:21509877

Anomalous diffusion measured by a twice-refocused spin echo pulse sequence: analysis using fractional order calculus.

PubMed

Gao, Qing; Srinivasan, Girish; Magin, Richard L; Zhou, Xiaohong Joe

2011-05-01

To theoretically develop and experimentally validate a formulism based on a fractional order calculus (FC) diffusion model to characterize anomalous diffusion in brain tissues measured with a twice-refocused spin-echo (TRSE) pulse sequence. The FC diffusion model is the fractional order generalization of the Bloch-Torrey equation. Using this model, an analytical expression was derived to describe the diffusion-induced signal attenuation in a TRSE pulse sequence. To experimentally validate this expression, a set of diffusion-weighted (DW) images was acquired at 3 Tesla from healthy human brains using a TRSE sequence with twelve b-values ranging from 0 to 2600 s/mm(2). For comparison, DW images were also acquired using a Stejskal-Tanner diffusion gradient in a single-shot spin-echo echo planar sequence. For both datasets, a Levenberg-Marquardt fitting algorithm was used to extract three parameters: diffusion coefficient D, fractional order derivative in space β, and a spatial parameter μ (in units of μm). Using adjusted R-squared values and standard deviations, D, β, and μ values and the goodness-of-fit in three specific regions of interest (ROIs) in white matter, gray matter, and cerebrospinal fluid, respectively, were evaluated for each of the two datasets. In addition, spatially resolved parametric maps were assessed qualitatively. The analytical expression for the TRSE sequence, derived from the FC diffusion model, accurately characterized the diffusion-induced signal loss in brain tissues at high b-values. In the selected ROIs, the goodness-of-fit and standard deviations for the TRSE dataset were comparable with the results obtained from the Stejskal-Tanner dataset, demonstrating the robustness of the FC model across multiple data acquisition strategies. Qualitatively, the D, β, and μ maps from the TRSE dataset exhibited fewer artifacts, reflecting the improved immunity to eddy currents. The diffusion-induced signal attenuation in a TRSE pulse sequence can be described by an FC diffusion model at high b-values. This model performs equally well for data acquired from the human brain tissues with a TRSE pulse sequence or a conventional Stejskal-Tanner sequence. Copyright © 2011 Wiley-Liss, Inc.

GOLabeler: Improving Sequence-based Large-scale Protein Function Prediction by Learning to Rank.

PubMed

You, Ronghui; Zhang, Zihan; Xiong, Yi; Sun, Fengzhu; Mamitsuka, Hiroshi; Zhu, Shanfeng

2018-03-07

Gene Ontology (GO) has been widely used to annotate functions of proteins and understand their biological roles. Currently only <1% of more than 70 million proteins in UniProtKB have experimental GO annotations, implying the strong necessity of automated function prediction (AFP) of proteins, where AFP is a hard multilabel classification problem due to one protein with a diverse number of GO terms. Most of these proteins have only sequences as input information, indicating the importance of sequence-based AFP (SAFP: sequences are the only input). Furthermore homology-based SAFP tools are competitive in AFP competitions, while they do not necessarily work well for so-called difficult proteins, which have <60% sequence identity to proteins with annotations already. Thus the vital and challenging problem now is how to develop a method for SAFP, particularly for difficult proteins. The key of this method is to extract not only homology information but also diverse, deep- rooted information/evidence from sequence inputs and integrate them into a predictor in a both effective and efficient manner. We propose GOLabeler, which integrates five component classifiers, trained from different features, including GO term frequency, sequence alignment, amino acid trigram, domains and motifs, and biophysical properties, etc., in the framework of learning to rank (LTR), a paradigm of machine learning, especially powerful for multilabel classification. The empirical results obtained by examining GOLabeler extensively and thoroughly by using large-scale datasets revealed numerous favorable aspects of GOLabeler, including significant performance advantage over state-of-the-art AFP methods. http://datamining-iip.fudan.edu.cn/golabeler. zhusf@fudan.edu.cn. Supplementary data are available at Bioinformatics online.

A Benchmark Dataset and Saliency-guided Stacked Autoencoders for Video-based Salient Object Detection.

PubMed

Li, Jia; Xia, Changqun; Chen, Xiaowu

2017-10-12

Image-based salient object detection (SOD) has been extensively studied in past decades. However, video-based SOD is much less explored due to the lack of large-scale video datasets within which salient objects are unambiguously defined and annotated. Toward this end, this paper proposes a video-based SOD dataset that consists of 200 videos. In constructing the dataset, we manually annotate all objects and regions over 7,650 uniformly sampled keyframes and collect the eye-tracking data of 23 subjects who free-view all videos. From the user data, we find that salient objects in a video can be defined as objects that consistently pop-out throughout the video, and objects with such attributes can be unambiguously annotated by combining manually annotated object/region masks with eye-tracking data of multiple subjects. To the best of our knowledge, it is currently the largest dataset for videobased salient object detection. Based on this dataset, this paper proposes an unsupervised baseline approach for video-based SOD by using saliencyguided stacked autoencoders. In the proposed approach, multiple spatiotemporal saliency cues are first extracted at the pixel, superpixel and object levels. With these saliency cues, stacked autoencoders are constructed in an unsupervised manner that automatically infers a saliency score for each pixel by progressively encoding the high-dimensional saliency cues gathered from the pixel and its spatiotemporal neighbors. In experiments, the proposed unsupervised approach is compared with 31 state-of-the-art models on the proposed dataset and outperforms 30 of them, including 19 imagebased classic (unsupervised or non-deep learning) models, six image-based deep learning models, and five video-based unsupervised models. Moreover, benchmarking results show that the proposed dataset is very challenging and has the potential to boost the development of video-based SOD.

TranslatomeDB: a comprehensive database and cloud-based analysis platform for translatome sequencing data

PubMed Central

Liu, Wanting; Xiang, Lunping; Zheng, Tingkai; Jin, Jingjie

2018-01-01

Abstract Translation is a key regulatory step, linking transcriptome and proteome. Two major methods of translatome investigations are RNC-seq (sequencing of translating mRNA) and Ribo-seq (ribosome profiling). To facilitate the investigation of translation, we built a comprehensive database TranslatomeDB (http://www.translatomedb.net/) which provides collection and integrated analysis of published and user-generated translatome sequencing data. The current version includes 2453 Ribo-seq, 10 RNC-seq and their 1394 corresponding mRNA-seq datasets in 13 species. The database emphasizes the analysis functions in addition to the dataset collections. Differential gene expression (DGE) analysis can be performed between any two datasets of same species and type, both on transcriptome and translatome levels. The translation indices translation ratios, elongation velocity index and translational efficiency can be calculated to quantitatively evaluate translational initiation efficiency and elongation velocity, respectively. All datasets were analyzed using a unified, robust, accurate and experimentally-verifiable pipeline based on the FANSe3 mapping algorithm and edgeR for DGE analyzes. TranslatomeDB also allows users to upload their own datasets and utilize the identical unified pipeline to analyze their data. We believe that our TranslatomeDB is a comprehensive platform and knowledgebase on translatome and proteome research, releasing the biologists from complex searching, analyzing and comparing huge sequencing data without needing local computational power. PMID:29106630

Using nearly full-genome HIV sequence data improves phylogeny reconstruction in a simulated epidemic

PubMed Central

Yebra, Gonzalo; Hodcroft, Emma B.; Ragonnet-Cronin, Manon L.; Pillay, Deenan; Brown, Andrew J. Leigh; Fraser, Christophe; Kellam, Paul; de Oliveira, Tulio; Dennis, Ann; Hoppe, Anne; Kityo, Cissy; Frampton, Dan; Ssemwanga, Deogratius; Tanser, Frank; Keshani, Jagoda; Lingappa, Jairam; Herbeck, Joshua; Wawer, Maria; Essex, Max; Cohen, Myron S.; Paton, Nicholas; Ratmann, Oliver; Kaleebu, Pontiano; Hayes, Richard; Fidler, Sarah; Quinn, Thomas; Novitsky, Vladimir; Haywards, Andrew; Nastouli, Eleni; Morris, Steven; Clark, Duncan; Kozlakidis, Zisis

2016-01-01

HIV molecular epidemiology studies analyse viral pol gene sequences due to their availability, but whole genome sequencing allows to use other genes. We aimed to determine what gene(s) provide(s) the best approximation to the real phylogeny by analysing a simulated epidemic (created as part of the PANGEA_HIV project) with a known transmission tree. We sub-sampled a simulated dataset of 4662 sequences into different combinations of genes (gag-pol-env, gag-pol, gag, pol, env and partial pol) and sampling depths (100%, 60%, 20% and 5%), generating 100 replicates for each case. We built maximum-likelihood trees for each combination using RAxML (GTR + Γ), and compared their topologies to the corresponding true tree’s using CompareTree. The accuracy of the trees was significantly proportional to the length of the sequences used, with the gag-pol-env datasets showing the best performance and gag and partial pol sequences showing the worst. The lowest sampling depths (20% and 5%) greatly reduced the accuracy of tree reconstruction and showed high variability among replicates, especially when using the shortest gene datasets. In conclusion, using longer sequences derived from nearly whole genomes will improve the reliability of phylogenetic reconstruction. With low sample coverage, results can be highly variable, particularly when based on short sequences. PMID:28008945

Using nearly full-genome HIV sequence data improves phylogeny reconstruction in a simulated epidemic.

PubMed

Yebra, Gonzalo; Hodcroft, Emma B; Ragonnet-Cronin, Manon L; Pillay, Deenan; Brown, Andrew J Leigh

2016-12-23

HIV molecular epidemiology studies analyse viral pol gene sequences due to their availability, but whole genome sequencing allows to use other genes. We aimed to determine what gene(s) provide(s) the best approximation to the real phylogeny by analysing a simulated epidemic (created as part of the PANGEA_HIV project) with a known transmission tree. We sub-sampled a simulated dataset of 4662 sequences into different combinations of genes (gag-pol-env, gag-pol, gag, pol, env and partial pol) and sampling depths (100%, 60%, 20% and 5%), generating 100 replicates for each case. We built maximum-likelihood trees for each combination using RAxML (GTR + Γ), and compared their topologies to the corresponding true tree's using CompareTree. The accuracy of the trees was significantly proportional to the length of the sequences used, with the gag-pol-env datasets showing the best performance and gag and partial pol sequences showing the worst. The lowest sampling depths (20% and 5%) greatly reduced the accuracy of tree reconstruction and showed high variability among replicates, especially when using the shortest gene datasets. In conclusion, using longer sequences derived from nearly whole genomes will improve the reliability of phylogenetic reconstruction. With low sample coverage, results can be highly variable, particularly when based on short sequences.

A phylogenetic overview of the antrodia clade (Basidiomycota, Polyporales)

Treesearch

Beatriz Ortiz-Santana; Daniel L. Lindner; Otto Miettinen; Alfredo Justo; David S. Hibbett

2013-01-01

Phylogenetic relationships among members of the antrodia clade were investigated with molecular data from two nuclear ribosomal DNA regions, LSU and ITS. A total of 123 species representing 26 genera producing a brown rot were included in the present study. Three DNA datasets (combined LSU-ITS dataset, LSU dataset, ITS dataset) comprising sequences of 449 isolates were...

DeepSynergy: predicting anti-cancer drug synergy with Deep Learning

PubMed Central

Preuer, Kristina; Lewis, Richard P I; Hochreiter, Sepp; Bender, Andreas; Bulusu, Krishna C; Klambauer, Günter

2018-01-01

Abstract Motivation While drug combination therapies are a well-established concept in cancer treatment, identifying novel synergistic combinations is challenging due to the size of combinatorial space. However, computational approaches have emerged as a time- and cost-efficient way to prioritize combinations to test, based on recently available large-scale combination screening data. Recently, Deep Learning has had an impact in many research areas by achieving new state-of-the-art model performance. However, Deep Learning has not yet been applied to drug synergy prediction, which is the approach we present here, termed DeepSynergy. DeepSynergy uses chemical and genomic information as input information, a normalization strategy to account for input data heterogeneity, and conical layers to model drug synergies. Results DeepSynergy was compared to other machine learning methods such as Gradient Boosting Machines, Random Forests, Support Vector Machines and Elastic Nets on the largest publicly available synergy dataset with respect to mean squared error. DeepSynergy significantly outperformed the other methods with an improvement of 7.2% over the second best method at the prediction of novel drug combinations within the space of explored drugs and cell lines. At this task, the mean Pearson correlation coefficient between the measured and the predicted values of DeepSynergy was 0.73. Applying DeepSynergy for classification of these novel drug combinations resulted in a high predictive performance of an AUC of 0.90. Furthermore, we found that all compared methods exhibit low predictive performance when extrapolating to unexplored drugs or cell lines, which we suggest is due to limitations in the size and diversity of the dataset. We envision that DeepSynergy could be a valuable tool for selecting novel synergistic drug combinations. Availability and implementation DeepSynergy is available via www.bioinf.jku.at/software/DeepSynergy. Contact klambauer@bioinf.jku.at Supplementary information Supplementary data are available at Bioinformatics online. PMID:29253077

Object recognition using deep convolutional neural networks with complete transfer and partial frozen layers

NASA Astrophysics Data System (ADS)

Kruithof, Maarten C.; Bouma, Henri; Fischer, Noëlle M.; Schutte, Klamer

2016-10-01

Object recognition is important to understand the content of video and allow flexible querying in a large number of cameras, especially for security applications. Recent benchmarks show that deep convolutional neural networks are excellent approaches for object recognition. This paper describes an approach of domain transfer, where features learned from a large annotated dataset are transferred to a target domain where less annotated examples are available as is typical for the security and defense domain. Many of these networks trained on natural images appear to learn features similar to Gabor filters and color blobs in the first layer. These first-layer features appear to be generic for many datasets and tasks while the last layer is specific. In this paper, we study the effect of copying all layers and fine-tuning a variable number. We performed an experiment with a Caffe-based network on 1000 ImageNet classes that are randomly divided in two equal subgroups for the transfer from one to the other. We copy all layers and vary the number of layers that is fine-tuned and the size of the target dataset. We performed additional experiments with the Keras platform on CIFAR-10 dataset to validate general applicability. We show with both platforms and both datasets that the accuracy on the target dataset improves when more target data is used. When the target dataset is large, it is beneficial to freeze only a few layers. For a large target dataset, the network without transfer learning performs better than the transfer network, especially if many layers are frozen. When the target dataset is small, it is beneficial to transfer (and freeze) many layers. For a small target dataset, the transfer network boosts generalization and it performs much better than the network without transfer learning. Learning time can be reduced by freezing many layers in a network.

BayesMotif: de novo protein sorting motif discovery from impure datasets.

PubMed

Hu, Jianjun; Zhang, Fan

2010-01-18

Protein sorting is the process that newly synthesized proteins are transported to their target locations within or outside of the cell. This process is precisely regulated by protein sorting signals in different forms. A major category of sorting signals are amino acid sub-sequences usually located at the N-terminals or C-terminals of protein sequences. Genome-wide experimental identification of protein sorting signals is extremely time-consuming and costly. Effective computational algorithms for de novo discovery of protein sorting signals is needed to improve the understanding of protein sorting mechanisms. We formulated the protein sorting motif discovery problem as a classification problem and proposed a Bayesian classifier based algorithm (BayesMotif) for de novo identification of a common type of protein sorting motifs in which a highly conserved anchor is present along with a less conserved motif regions. A false positive removal procedure is developed to iteratively remove sequences that are unlikely to contain true motifs so that the algorithm can identify motifs from impure input sequences. Experiments on both implanted motif datasets and real-world datasets showed that the enhanced BayesMotif algorithm can identify anchored sorting motifs from pure or impure protein sequence dataset. It also shows that the false positive removal procedure can help to identify true motifs even when there is only 20% of the input sequences containing true motif instances. We proposed BayesMotif, a novel Bayesian classification based algorithm for de novo discovery of a special category of anchored protein sorting motifs from impure datasets. Compared to conventional motif discovery algorithms such as MEME, our algorithm can find less-conserved motifs with short highly conserved anchors. Our algorithm also has the advantage of easy incorporation of additional meta-sequence features such as hydrophobicity or charge of the motifs which may help to overcome the limitations of PWM (position weight matrix) motif model.

Pooled assembly of marine metagenomic datasets: enriching annotation through chimerism.

PubMed

Magasin, Jonathan D; Gerloff, Dietlind L

2015-02-01

Despite advances in high-throughput sequencing, marine metagenomic samples remain largely opaque. A typical sample contains billions of microbial organisms from thousands of genomes and quadrillions of DNA base pairs. Its derived metagenomic dataset underrepresents this complexity by orders of magnitude because of the sparseness and shortness of sequencing reads. Read shortness and sequencing errors pose a major challenge to accurate species and functional annotation. This includes distinguishing known from novel species. Often the majority of reads cannot be annotated and thus cannot help our interpretation of the sample. Here, we demonstrate quantitatively how careful assembly of marine metagenomic reads within, but also across, datasets can alleviate this problem. For 10 simulated datasets, each with species complexity modeled on a real counterpart, chimerism remained within the same species for most contigs (97%). For 42 real pyrosequencing ('454') datasets, assembly increased the proportion of annotated reads, and even more so when datasets were pooled, by on average 1.6% (max 6.6%) for species, 9.0% (max 28.7%) for Pfam protein domains and 9.4% (max 22.9%) for PANTHER gene families. Our results outline exciting prospects for data sharing in the metagenomics community. While chimeric sequences should be avoided in other areas of metagenomics (e.g. biodiversity analyses), conservative pooled assembly is advantageous for annotation specificity and sensitivity. Intriguingly, our experiment also found potential prospects for (low-cost) discovery of new species in 'old' data. dgerloff@ffame.org Supplementary data are available at Bioinformatics online. © The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.

The Sorcerer II Global Ocean Sampling expedition: expanding the universe of protein families.

PubMed

Yooseph, Shibu; Sutton, Granger; Rusch, Douglas B; Halpern, Aaron L; Williamson, Shannon J; Remington, Karin; Eisen, Jonathan A; Heidelberg, Karla B; Manning, Gerard; Li, Weizhong; Jaroszewski, Lukasz; Cieplak, Piotr; Miller, Christopher S; Li, Huiying; Mashiyama, Susan T; Joachimiak, Marcin P; van Belle, Christopher; Chandonia, John-Marc; Soergel, David A; Zhai, Yufeng; Natarajan, Kannan; Lee, Shaun; Raphael, Benjamin J; Bafna, Vineet; Friedman, Robert; Brenner, Steven E; Godzik, Adam; Eisenberg, David; Dixon, Jack E; Taylor, Susan S; Strausberg, Robert L; Frazier, Marvin; Venter, J Craig

2007-03-01

Metagenomics projects based on shotgun sequencing of populations of micro-organisms yield insight into protein families. We used sequence similarity clustering to explore proteins with a comprehensive dataset consisting of sequences from available databases together with 6.12 million proteins predicted from an assembly of 7.7 million Global Ocean Sampling (GOS) sequences. The GOS dataset covers nearly all known prokaryotic protein families. A total of 3,995 medium- and large-sized clusters consisting of only GOS sequences are identified, out of which 1,700 have no detectable homology to known families. The GOS-only clusters contain a higher than expected proportion of sequences of viral origin, thus reflecting a poor sampling of viral diversity until now. Protein domain distributions in the GOS dataset and current protein databases show distinct biases. Several protein domains that were previously categorized as kingdom specific are shown to have GOS examples in other kingdoms. About 6,000 sequences (ORFans) from the literature that heretofore lacked similarity to known proteins have matches in the GOS data. The GOS dataset is also used to improve remote homology detection. Overall, besides nearly doubling the number of current proteins, the predicted GOS proteins also add a great deal of diversity to known protein families and shed light on their evolution. These observations are illustrated using several protein families, including phosphatases, proteases, ultraviolet-irradiation DNA damage repair enzymes, glutamine synthetase, and RuBisCO. The diversity added by GOS data has implications for choosing targets for experimental structure characterization as part of structural genomics efforts. Our analysis indicates that new families are being discovered at a rate that is linear or almost linear with the addition of new sequences, implying that we are still far from discovering all protein families in nature.

A Deep Learning Approach to on-Node Sensor Data Analytics for Mobile or Wearable Devices.

PubMed

Ravi, Daniele; Wong, Charence; Lo, Benny; Yang, Guang-Zhong

2017-01-01

The increasing popularity of wearable devices in recent years means that a diverse range of physiological and functional data can now be captured continuously for applications in sports, wellbeing, and healthcare. This wealth of information requires efficient methods of classification and analysis where deep learning is a promising technique for large-scale data analytics. While deep learning has been successful in implementations that utilize high-performance computing platforms, its use on low-power wearable devices is limited by resource constraints. In this paper, we propose a deep learning methodology, which combines features learned from inertial sensor data together with complementary information from a set of shallow features to enable accurate and real-time activity classification. The design of this combined method aims to overcome some of the limitations present in a typical deep learning framework where on-node computation is required. To optimize the proposed method for real-time on-node computation, spectral domain preprocessing is used before the data are passed onto the deep learning framework. The classification accuracy of our proposed deep learning approach is evaluated against state-of-the-art methods using both laboratory and real world activity datasets. Our results show the validity of the approach on different human activity datasets, outperforming other methods, including the two methods used within our combined pipeline. We also demonstrate that the computation times for the proposed method are consistent with the constraints of real-time on-node processing on smartphones and a wearable sensor platform.

Population-genomic variation within RNA viruses of the Western honey bee, Apis mellifera, inferred from deep sequencing

USDA-ARS?s Scientific Manuscript database

Deep sequencing of viruses isolated from infected hosts is an efficient way to measure population-genetic variation and can reveal patterns of dispersal and natural selection. In this study, we mined existing Illumina sequence reads to investigate single-nucleotide polymorphisms (SNPs) within two RN...

DNA barcoding for species identification in deep-sea clams (Mollusca: Bivalvia: Vesicomyidae).

PubMed

Liu, Jun; Zhang, Haibin

2018-01-15

Deep-sea clams (Bivalvia: Vesicomyidae) have been found in reduced environments over the world oceans, but taxonomy of this group remains confusing at species and supraspecific levels due to their high-morphological similarity and plasticity. In the present study, we collected mitochondrial COI sequences to evaluate the utility of DNA barcoding on identifying vesicomyid species. COI dataset identified 56 well-supported putative species/operational taxonomic units (OTUs), approximately covering half of the extant vesicomyid species. One species (OTU2) was first detected, and may represent a new species. Average distances between species ranged from 1.65 to 29.64%, generally higher than average intraspecific distances (0-1.41%) when excluding Pliocardia sp.10 cf. venusta (average intraspecific distance 1.91%). Local barcoding gap existed in 33 of the 35 species when comparing distances of maximum interspecific and minimum interspecific distances with two exceptions (Abyssogena southwardae and Calyptogena rectimargo-starobogatovi). The barcode index number (BIN) system determined 41 of the 56 species/OTUs, each with a unique BIN, indicating their validity. Three species were found to have two BINs, together with their high level of intraspecific variation, implying cryptic diversity within them. Although fewer 16 S sequences were collected, similar results were obtained. Nineteen putative species were determined and no overlap observed between intra- and inter-specific variation. Implications of DNA barcoding for the Vesicomyidae taxonomy were then discussed. Findings of this study will provide important evidence for taxonomic revision in this problematic clam group, and accelerate the discovery of new vesicomyid species in the future.

Transcriptomic analysis and mutational status of IDH1 in paired primary-recurrent intrahepatic cholangiocarcinoma.

PubMed

Peraldo-Neia, C; Ostano, P; Cavalloni, G; Pignochino, Y; Sangiolo, D; De Cecco, L; Marchesi, E; Ribero, D; Scarpa, A; De Rose, A M; Giuliani, A; Calise, F; Raggi, C; Invernizzi, P; Aglietta, M; Chiorino, G; Leone, F

2018-06-05

Effective target therapies for intrahepatic cholangiocarcinoma (ICC) have not been identified so far. One of the reasons may be the genetic evolution from primary (PR) to recurrent (REC) tumors. We aim to identify peculiar characteristics and to select potential targets specific for recurrent tumors. Eighteen ICC paired PR and REC tumors were collected from 5 Italian Centers. Eleven pairs were analyzed for gene expression profiling and 16 for mutational status of IDH1. For one pair, deep mutational analysis by Next Generation Sequencing was also carried out. An independent cohort of patients was used for validation. Two class-paired comparison yielded 315 differentially expressed genes between REC and PR tumors. Up-regulated genes in RECs are involved in RNA/DNA processing, cell cycle, epithelial to mesenchymal transition (EMT), resistance to apoptosis, and cytoskeleton remodeling. Down-regulated genes participate to epithelial cell differentiation, proteolysis, apoptotic, immune response, and inflammatory processes. A 24 gene signature is able to discriminate RECs from PRs in an independent cohort; FANCG is statistically associated with survival in the chol-TCGA dataset. IDH1 was mutated in the RECs of five patients; 4 of them displayed the mutation only in RECs. Deep sequencing performed in one patient confirmed the IDH1 mutation in REC. RECs are enriched for genes involved in EMT, resistance to apoptosis, and cytoskeleton remodeling. Key players of these pathways might be considered druggable targets in RECs. IDH1 is mutated in 30% of RECs, becoming both a marker of progression and a target for therapy.

Deep learning for single-molecule science

NASA Astrophysics Data System (ADS)

Albrecht, Tim; Slabaugh, Gregory; Alonso, Eduardo; Al-Arif, SM Masudur R.

2017-10-01

Exploring and making predictions based on single-molecule data can be challenging, not only due to the sheer size of the datasets, but also because a priori knowledge about the signal characteristics is typically limited and poor signal-to-noise ratio. For example, hypothesis-driven data exploration, informed by an expectation of the signal characteristics, can lead to interpretation bias or loss of information. Equally, even when the different data categories are known, e.g., the four bases in DNA sequencing, it is often difficult to know how to make best use of the available information content. The latest developments in machine learning (ML), so-called deep learning (DL) offer interesting, new avenues to address such challenges. In some applications, such as speech and image recognition, DL has been able to outperform conventional ML strategies and even human performance. However, to date DL has not been applied much in single-molecule science, presumably in part because relatively little is known about the ‘internal workings’ of such DL tools within single-molecule science as a field. In this Tutorial, we make an attempt to illustrate in a step-by-step guide how one of those, a convolutional neural network (CNN), may be used for base calling in DNA sequencing applications. We compare it with a SVM as a more conventional ML method, and discuss some of the strengths and weaknesses of the approach. In particular, a ‘deep’ neural network has many features of a ‘black box’, which has important implications on how we look at and interpret data.

Detection of Emerging Vaccine-Related Polioviruses by Deep Sequencing.

PubMed

Sahoo, Malaya K; Holubar, Marisa; Huang, ChunHong; Mohamed-Hadley, Alisha; Liu, Yuanyuan; Waggoner, Jesse J; Troy, Stephanie B; Garcia-Garcia, Lourdes; Ferreyra-Reyes, Leticia; Maldonado, Yvonne; Pinsky, Benjamin A

2017-07-01

Oral poliovirus vaccine can mutate to regain neurovirulence. To date, evaluation of these mutations has been performed primarily on culture-enriched isolates by using conventional Sanger sequencing. We therefore developed a culture-independent, deep-sequencing method targeting the 5' untranslated region (UTR) and P1 genomic region to characterize vaccine-related poliovirus variants. Error analysis of the deep-sequencing method demonstrated reliable detection of poliovirus mutations at levels of <1%, depending on read depth. Sequencing of viral nucleic acids from the stool of vaccinated, asymptomatic children and their close contacts collected during a prospective cohort study in Veracruz, Mexico, revealed no vaccine-derived polioviruses. This was expected given that the longest duration between sequenced sample collection and the end of the most recent national immunization week was 66 days. However, we identified many low-level variants (<5%) distributed across the 5' UTR and P1 genomic region in all three Sabin serotypes, as well as vaccine-related viruses with multiple canonical mutations associated with phenotypic reversion present at high levels (>90%). These results suggest that monitoring emerging vaccine-related poliovirus variants by deep sequencing may aid in the poliovirus endgame and efforts to ensure global polio eradication. Copyright © 2017 Sahoo et al.

Annotation of the Transcriptome from Taenia pisiformis and Its Comparative Analysis with Three Taeniidae Species

PubMed Central

Yang, Deying; Fu, Yan; Wu, Xuhang; Xie, Yue; Nie, Huaming; Chen, Lin; Nong, Xiang; Gu, Xiaobin; Wang, Shuxian; Peng, Xuerong; Yan, Ning; Zhang, Runhui; Zheng, Wanpeng; Yang, Guangyou

2012-01-01

Background Taenia pisiformis is one of the most common intestinal tapeworms and can cause infections in canines. Adult T. pisiformis (canines as definitive hosts) and Cysticercus pisiformis (rabbits as intermediate hosts) cause significant health problems to the host and considerable socio-economic losses as a consequence. No complete genomic data regarding T. pisiformis are currently available in public databases. RNA-seq provides an effective approach to analyze the eukaryotic transcriptome to generate large functional gene datasets that can be used for further studies. Methodology/Principal Findings In this study, 2.67 million sequencing clean reads and 72,957 unigenes were generated using the RNA-seq technique. Based on a sequence similarity search with known proteins, a total of 26,012 unigenes (no redundancy) were identified after quality control procedures via the alignment of four databases. Overall, 15,920 unigenes were mapped to 203 Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. Through analyzing the glycolysis/gluconeogenesis and axonal guidance pathways, we achieved an in-depth understanding of the biochemistry of T. pisiformis. Here, we selected four unigenes at random and obtained their full-length cDNA clones using RACE PCR. Functional distribution characteristics were gained through comparing four cestode species (72,957 unigenes of T. pisiformis, 30,700 ESTs of T. solium, 1,058 ESTs of Eg+Em [conserved ESTs between Echinococcus granulosus and Echinococcus multilocularis]), with the cluster of orthologous groups (COG) and gene ontology (GO) functional classification systems. Furthermore, the conserved common genes in these four cestode species were obtained and aligned by the KEGG database. Conclusion This study provides an extensive transcriptome dataset obtained from the deep sequencing of T. pisiformis in a non-model whole genome. The identification of conserved genes may provide novel approaches for potential drug targets and vaccinations against cestode infections. Research can now accelerate into the functional genomics, immunity and gene expression profiles of cestode species. PMID:22514598

Feature Representations for Neuromorphic Audio Spike Streams.

PubMed

Anumula, Jithendar; Neil, Daniel; Delbruck, Tobi; Liu, Shih-Chii

2018-01-01

Event-driven neuromorphic spiking sensors such as the silicon retina and the silicon cochlea encode the external sensory stimuli as asynchronous streams of spikes across different channels or pixels. Combining state-of-art deep neural networks with the asynchronous outputs of these sensors has produced encouraging results on some datasets but remains challenging. While the lack of effective spiking networks to process the spike streams is one reason, the other reason is that the pre-processing methods required to convert the spike streams to frame-based features needed for the deep networks still require further investigation. This work investigates the effectiveness of synchronous and asynchronous frame-based features generated using spike count and constant event binning in combination with the use of a recurrent neural network for solving a classification task using N-TIDIGITS18 dataset. This spike-based dataset consists of recordings from the Dynamic Audio Sensor, a spiking silicon cochlea sensor, in response to the TIDIGITS audio dataset. We also propose a new pre-processing method which applies an exponential kernel on the output cochlea spikes so that the interspike timing information is better preserved. The results from the N-TIDIGITS18 dataset show that the exponential features perform better than the spike count features, with over 91% accuracy on the digit classification task. This accuracy corresponds to an improvement of at least 2.5% over the use of spike count features, establishing a new state of the art for this dataset.

Feature Representations for Neuromorphic Audio Spike Streams

PubMed Central

Anumula, Jithendar; Neil, Daniel; Delbruck, Tobi; Liu, Shih-Chii

2018-01-01

Event-driven neuromorphic spiking sensors such as the silicon retina and the silicon cochlea encode the external sensory stimuli as asynchronous streams of spikes across different channels or pixels. Combining state-of-art deep neural networks with the asynchronous outputs of these sensors has produced encouraging results on some datasets but remains challenging. While the lack of effective spiking networks to process the spike streams is one reason, the other reason is that the pre-processing methods required to convert the spike streams to frame-based features needed for the deep networks still require further investigation. This work investigates the effectiveness of synchronous and asynchronous frame-based features generated using spike count and constant event binning in combination with the use of a recurrent neural network for solving a classification task using N-TIDIGITS18 dataset. This spike-based dataset consists of recordings from the Dynamic Audio Sensor, a spiking silicon cochlea sensor, in response to the TIDIGITS audio dataset. We also propose a new pre-processing method which applies an exponential kernel on the output cochlea spikes so that the interspike timing information is better preserved. The results from the N-TIDIGITS18 dataset show that the exponential features perform better than the spike count features, with over 91% accuracy on the digit classification task. This accuracy corresponds to an improvement of at least 2.5% over the use of spike count features, establishing a new state of the art for this dataset. PMID:29479300

Rational Protein Engineering Guided by Deep Mutational Scanning

PubMed Central

Shin, HyeonSeok; Cho, Byung-Kwan

2015-01-01

Sequence–function relationship in a protein is commonly determined by the three-dimensional protein structure followed by various biochemical experiments. However, with the explosive increase in the number of genome sequences, facilitated by recent advances in sequencing technology, the gap between protein sequences available and three-dimensional structures is rapidly widening. A recently developed method termed deep mutational scanning explores the functional phenotype of thousands of mutants via massive sequencing. Coupled with a highly efficient screening system, this approach assesses the phenotypic changes made by the substitution of each amino acid sequence that constitutes a protein. Such an informational resource provides the functional role of each amino acid sequence, thereby providing sufficient rationale for selecting target residues for protein engineering. Here, we discuss the current applications of deep mutational scanning and consider experimental design. PMID:26404267

Deep Sequencing in Infectious Diseases: Immune and Pathogen Repertoires for the Improvement of Patient Outcomes.

PubMed

Burkholder, William F; Newell, Evan W; Poidinger, Michael; Chen, Swaine; Fink, Katja

2017-01-01

The inaugural workshop "Deep Sequencing in Infectious Diseases: Immune and Pathogen Repertoires for the Improvement of Patient Outcomes" was held in Singapore on 13-14 October 2016. The aim of the workshop was to discuss the latest trends in using high-throughput sequencing, bioinformatics, and allied technologies to analyze immune and pathogen repertoires and their interplay within the host, bringing together key international players in the field and Singapore-based researchers and clinician-scientists. The focus was in particular on the application of these technologies for the improvement of patient diagnosis, prognosis and treatment, and for other broad public health outcomes. The presentations by scientists and clinicians showed the potential of deep sequencing technology to capture the coevolution of adaptive immunity and pathogens. For clinical applications, some key challenges remain, such as the long turnaround time and relatively high cost of deep sequencing for pathogen identification and characterization and the lack of international standardization in immune repertoire analysis.

Deep Sequencing in Infectious Diseases: Immune and Pathogen Repertoires for the Improvement of Patient Outcomes

PubMed Central

Burkholder, William F.; Newell, Evan W.; Poidinger, Michael; Chen, Swaine; Fink, Katja

2017-01-01

The inaugural workshop “Deep Sequencing in Infectious Diseases: Immune and Pathogen Repertoires for the Improvement of Patient Outcomes” was held in Singapore on 13–14 October 2016. The aim of the workshop was to discuss the latest trends in using high-throughput sequencing, bioinformatics, and allied technologies to analyze immune and pathogen repertoires and their interplay within the host, bringing together key international players in the field and Singapore-based researchers and clinician-scientists. The focus was in particular on the application of these technologies for the improvement of patient diagnosis, prognosis and treatment, and for other broad public health outcomes. The presentations by scientists and clinicians showed the potential of deep sequencing technology to capture the coevolution of adaptive immunity and pathogens. For clinical applications, some key challenges remain, such as the long turnaround time and relatively high cost of deep sequencing for pathogen identification and characterization and the lack of international standardization in immune repertoire analysis. PMID:28620372

Representation learning: a unified deep learning framework for automatic prostate MR segmentation.

PubMed

Liao, Shu; Gao, Yaozong; Oto, Aytekin; Shen, Dinggang

2013-01-01

Image representation plays an important role in medical image analysis. The key to the success of different medical image analysis algorithms is heavily dependent on how we represent the input data, namely features used to characterize the input image. In the literature, feature engineering remains as an active research topic, and many novel hand-crafted features are designed such as Haar wavelet, histogram of oriented gradient, and local binary patterns. However, such features are not designed with the guidance of the underlying dataset at hand. To this end, we argue that the most effective features should be designed in a learning based manner, namely representation learning, which can be adapted to different patient datasets at hand. In this paper, we introduce a deep learning framework to achieve this goal. Specifically, a stacked independent subspace analysis (ISA) network is adopted to learn the most effective features in a hierarchical and unsupervised manner. The learnt features are adapted to the dataset at hand and encode high level semantic anatomical information. The proposed method is evaluated on the application of automatic prostate MR segmentation. Experimental results show that significant segmentation accuracy improvement can be achieved by the proposed deep learning method compared to other state-of-the-art segmentation approaches.

Multiview human activity recognition system based on spatiotemporal template for video surveillance system

NASA Astrophysics Data System (ADS)

Kushwaha, Alok Kumar Singh; Srivastava, Rajeev

2015-09-01

An efficient view invariant framework for the recognition of human activities from an input video sequence is presented. The proposed framework is composed of three consecutive modules: (i) detect and locate people by background subtraction, (ii) view invariant spatiotemporal template creation for different activities, (iii) and finally, template matching is performed for view invariant activity recognition. The foreground objects present in a scene are extracted using change detection and background modeling. The view invariant templates are constructed using the motion history images and object shape information for different human activities in a video sequence. For matching the spatiotemporal templates for various activities, the moment invariants and Mahalanobis distance are used. The proposed approach is tested successfully on our own viewpoint dataset, KTH action recognition dataset, i3DPost multiview dataset, MSR viewpoint action dataset, VideoWeb multiview dataset, and WVU multiview human action recognition dataset. From the experimental results and analysis over the chosen datasets, it is observed that the proposed framework is robust, flexible, and efficient with respect to multiple views activity recognition, scale, and phase variations.

SEXCMD: Development and validation of sex marker sequences for whole-exome/genome and RNA sequencing.

PubMed

Jeong, Seongmun; Kim, Jiwoong; Park, Won; Jeon, Hongmin; Kim, Namshin

2017-01-01

Over the last decade, a large number of nucleotide sequences have been generated by next-generation sequencing technologies and deposited to public databases. However, most of these datasets do not specify the sex of individuals sampled because researchers typically ignore or hide this information. Male and female genomes in many species have distinctive sex chromosomes, XX/XY and ZW/ZZ, and expression levels of many sex-related genes differ between the sexes. Herein, we describe how to develop sex marker sequences from syntenic regions of sex chromosomes and use them to quickly identify the sex of individuals being analyzed. Array-based technologies routinely use either known sex markers or the B-allele frequency of X or Z chromosomes to deduce the sex of an individual. The same strategy has been used with whole-exome/genome sequence data; however, all reads must be aligned onto a reference genome to determine the B-allele frequency of the X or Z chromosomes. SEXCMD is a pipeline that can extract sex marker sequences from reference sex chromosomes and rapidly identify the sex of individuals from whole-exome/genome and RNA sequencing after training with a known dataset through a simple machine learning approach. The pipeline counts total numbers of hits from sex-specific marker sequences and identifies the sex of the individuals sampled based on the fact that XX/ZZ samples do not have Y or W chromosome hits. We have successfully validated our pipeline with mammalian (Homo sapiens; XY) and avian (Gallus gallus; ZW) genomes. Typical calculation time when applying SEXCMD to human whole-exome or RNA sequencing datasets is a few minutes, and analyzing human whole-genome datasets takes about 10 minutes. Another important application of SEXCMD is as a quality control measure to avoid mixing samples before bioinformatics analysis. SEXCMD comprises simple Python and R scripts and is freely available at https://github.com/lovemun/SEXCMD.

SACCHARIS: an automated pipeline to streamline discovery of carbohydrate active enzyme activities within polyspecific families and de novo sequence datasets.

PubMed

Jones, Darryl R; Thomas, Dallas; Alger, Nicholas; Ghavidel, Ata; Inglis, G Douglas; Abbott, D Wade

2018-01-01

Deposition of new genetic sequences in online databases is expanding at an unprecedented rate. As a result, sequence identification continues to outpace functional characterization of carbohydrate active enzymes (CAZymes). In this paradigm, the discovery of enzymes with novel functions is often hindered by high volumes of uncharacterized sequences particularly when the enzyme sequence belongs to a family that exhibits diverse functional specificities (i.e., polyspecificity). Therefore, to direct sequence-based discovery and characterization of new enzyme activities we have developed an automated in silico pipeline entitled: Sequence Analysis and Clustering of CarboHydrate Active enzymes for Rapid Informed prediction of Specificity (SACCHARIS). This pipeline streamlines the selection of uncharacterized sequences for discovery of new CAZyme or CBM specificity from families currently maintained on the CAZy website or within user-defined datasets. SACCHARIS was used to generate a phylogenetic tree of a GH43, a CAZyme family with defined subfamily designations. This analysis confirmed that large datasets can be organized into sequence clusters of manageable sizes that possess related functions. Seeding this tree with a GH43 sequence from Bacteroides dorei DSM 17855 (BdGH43b, revealed it partitioned as a single sequence within the tree. This pattern was consistent with it possessing a unique enzyme activity for GH43 as BdGH43b is the first described α-glucanase described for this family. The capacity of SACCHARIS to extract and cluster characterized carbohydrate binding module sequences was demonstrated using family 6 CBMs (i.e., CBM6s). This CBM family displays a polyspecific ligand binding profile and contains many structurally determined members. Using SACCHARIS to identify a cluster of divergent sequences, a CBM6 sequence from a unique clade was demonstrated to bind yeast mannan, which represents the first description of an α-mannan binding CBM. Additionally, we have performed a CAZome analysis of an in-house sequenced bacterial genome and a comparative analysis of B. thetaiotaomicron VPI-5482 and B. thetaiotaomicron 7330, to demonstrate that SACCHARIS can generate "CAZome fingerprints", which differentiate between the saccharolytic potential of two related strains in silico. Establishing sequence-function and sequence-structure relationships in polyspecific CAZyme families are promising approaches for streamlining enzyme discovery. SACCHARIS facilitates this process by embedding CAZyme and CBM family trees generated from biochemically to structurally characterized sequences, with protein sequences that have unknown functions. In addition, these trees can be integrated with user-defined datasets (e.g., genomics, metagenomics, and transcriptomics) to inform experimental characterization of new CAZymes or CBMs not currently curated, and for researchers to compare differential sequence patterns between entire CAZomes. In this light, SACCHARIS provides an in silico tool that can be tailored for enzyme bioprospecting in datasets of increasing complexity and for diverse applications in glycobiotechnology.

BCDForest: a boosting cascade deep forest model towards the classification of cancer subtypes based on gene expression data.

PubMed

Guo, Yang; Liu, Shuhui; Li, Zhanhuai; Shang, Xuequn

2018-04-11

The classification of cancer subtypes is of great importance to cancer disease diagnosis and therapy. Many supervised learning approaches have been applied to cancer subtype classification in the past few years, especially of deep learning based approaches. Recently, the deep forest model has been proposed as an alternative of deep neural networks to learn hyper-representations by using cascade ensemble decision trees. It has been proved that the deep forest model has competitive or even better performance than deep neural networks in some extent. However, the standard deep forest model may face overfitting and ensemble diversity challenges when dealing with small sample size and high-dimensional biology data. In this paper, we propose a deep learning model, so-called BCDForest, to address cancer subtype classification on small-scale biology datasets, which can be viewed as a modification of the standard deep forest model. The BCDForest distinguishes from the standard deep forest model with the following two main contributions: First, a named multi-class-grained scanning method is proposed to train multiple binary classifiers to encourage diversity of ensemble. Meanwhile, the fitting quality of each classifier is considered in representation learning. Second, we propose a boosting strategy to emphasize more important features in cascade forests, thus to propagate the benefits of discriminative features among cascade layers to improve the classification performance. Systematic comparison experiments on both microarray and RNA-Seq gene expression datasets demonstrate that our method consistently outperforms the state-of-the-art methods in application of cancer subtype classification. The multi-class-grained scanning and boosting strategy in our model provide an effective solution to ease the overfitting challenge and improve the robustness of deep forest model working on small-scale data. Our model provides a useful approach to the classification of cancer subtypes by using deep learning on high-dimensional and small-scale biology data.

Deep neural networks for texture classification-A theoretical analysis.

PubMed

Basu, Saikat; Mukhopadhyay, Supratik; Karki, Manohar; DiBiano, Robert; Ganguly, Sangram; Nemani, Ramakrishna; Gayaka, Shreekant

2018-01-01

We investigate the use of Deep Neural Networks for the classification of image datasets where texture features are important for generating class-conditional discriminative representations. To this end, we first derive the size of the feature space for some standard textural features extracted from the input dataset and then use the theory of Vapnik-Chervonenkis dimension to show that hand-crafted feature extraction creates low-dimensional representations which help in reducing the overall excess error rate. As a corollary to this analysis, we derive for the first time upper bounds on the VC dimension of Convolutional Neural Network as well as Dropout and Dropconnect networks and the relation between excess error rate of Dropout and Dropconnect networks. The concept of intrinsic dimension is used to validate the intuition that texture-based datasets are inherently higher dimensional as compared to handwritten digits or other object recognition datasets and hence more difficult to be shattered by neural networks. We then derive the mean distance from the centroid to the nearest and farthest sampling points in an n-dimensional manifold and show that the Relative Contrast of the sample data vanishes as dimensionality of the underlying vector space tends to infinity. Copyright © 2017 Elsevier Ltd. All rights reserved.

Deep learning based beat event detection in action movie franchises

NASA Astrophysics Data System (ADS)

Ejaz, N.; Khan, U. A.; Martínez-del-Amor, M. A.; Sparenberg, H.

2018-04-01

Automatic understanding and interpretation of movies can be used in a variety of ways to semantically manage the massive volumes of movies data. "Action Movie Franchises" dataset is a collection of twenty Hollywood action movies from five famous franchises with ground truth annotations at shot and beat level of each movie. In this dataset, the annotations are provided for eleven semantic beat categories. In this work, we propose a deep learning based method to classify shots and beat-events on this dataset. The training dataset for each of the eleven beat categories is developed and then a Convolution Neural Network is trained. After finding the shot boundaries, key frames are extracted for each shot and then three classification labels are assigned to each key frame. The classification labels for each of the key frames in a particular shot are then used to assign a unique label to each shot. A simple sliding window based method is then used to group adjacent shots having the same label in order to find a particular beat event. The results of beat event classification are presented based on criteria of precision, recall, and F-measure. The results are compared with the existing technique and significant improvements are recorded.

Multiclass classification of obstructive sleep apnea/hypopnea based on a convolutional neural network from a single-lead electrocardiogram.

PubMed

Urtnasan, Erdenebayar; Park, Jong-Uk; Lee, Kyoung-Joung

2018-05-24

In this paper, we propose a convolutional neural network (CNN)-based deep learning architecture for multiclass classification of obstructive sleep apnea and hypopnea (OSAH) using single-lead electrocardiogram (ECG) recordings. OSAH is the most common sleep-related breathing disorder. Many subjects who suffer from OSAH remain undiagnosed; thus, early detection of OSAH is important. In this study, automatic classification of three classes-normal, hypopnea, and apnea-based on a CNN is performed. An optimal six-layer CNN model is trained on a training dataset (45,096 events) and evaluated on a test dataset (11,274 events). The training set (69 subjects) and test set (17 subjects) were collected from 86 subjects with length of approximately 6 h and segmented into 10 s durations. The proposed CNN model reaches a mean -score of 93.0 for the training dataset and 87.0 for the test dataset. Thus, proposed deep learning architecture achieved a high performance for multiclass classification of OSAH using single-lead ECG recordings. The proposed method can be employed in screening of patients suspected of having OSAH. © 2018 Institute of Physics and Engineering in Medicine.

Fast and Sensitive Alignment of Microbial Whole Genome Sequencing Reads to Large Sequence Datasets on a Desktop PC: Application to Metagenomic Datasets and Pathogen Identification

PubMed Central

2014-01-01

Next generation sequencing (NGS) of metagenomic samples is becoming a standard approach to detect individual species or pathogenic strains of microorganisms. Computer programs used in the NGS community have to balance between speed and sensitivity and as a result, species or strain level identification is often inaccurate and low abundance pathogens can sometimes be missed. We have developed Taxoner, an open source, taxon assignment pipeline that includes a fast aligner (e.g. Bowtie2) and a comprehensive DNA sequence database. We tested the program on simulated datasets as well as experimental data from Illumina, IonTorrent, and Roche 454 sequencing platforms. We found that Taxoner performs as well as, and often better than BLAST, but requires two orders of magnitude less running time meaning that it can be run on desktop or laptop computers. Taxoner is slower than the approaches that use small marker databases but is more sensitive due the comprehensive reference database. In addition, it can be easily tuned to specific applications using small tailored databases. When applied to metagenomic datasets, Taxoner can provide a functional summary of the genes mapped and can provide strain level identification. Taxoner is written in C for Linux operating systems. The code and documentation are available for research applications at http://code.google.com/p/taxoner. PMID:25077800

Characterization and prediction of residues determining protein functional specificity.

PubMed

Capra, John A; Singh, Mona

2008-07-01

Within a homologous protein family, proteins may be grouped into subtypes that share specific functions that are not common to the entire family. Often, the amino acids present in a small number of sequence positions determine each protein's particular functional specificity. Knowledge of these specificity determining positions (SDPs) aids in protein function prediction, drug design and experimental analysis. A number of sequence-based computational methods have been introduced for identifying SDPs; however, their further development and evaluation have been hindered by the limited number of known experimentally determined SDPs. We combine several bioinformatics resources to automate a process, typically undertaken manually, to build a dataset of SDPs. The resulting large dataset, which consists of SDPs in enzymes, enables us to characterize SDPs in terms of their physicochemical and evolutionary properties. It also facilitates the large-scale evaluation of sequence-based SDP prediction methods. We present a simple sequence-based SDP prediction method, GroupSim, and show that, surprisingly, it is competitive with a representative set of current methods. We also describe ConsWin, a heuristic that considers sequence conservation of neighboring amino acids, and demonstrate that it improves the performance of all methods tested on our large dataset of enzyme SDPs. Datasets and GroupSim code are available online at http://compbio.cs.princeton.edu/specificity/. Supplementary data are available at Bioinformatics online.

Fast and sensitive alignment of microbial whole genome sequencing reads to large sequence datasets on a desktop PC: application to metagenomic datasets and pathogen identification.

PubMed

Pongor, Lőrinc S; Vera, Roberto; Ligeti, Balázs

2014-01-01

Next generation sequencing (NGS) of metagenomic samples is becoming a standard approach to detect individual species or pathogenic strains of microorganisms. Computer programs used in the NGS community have to balance between speed and sensitivity and as a result, species or strain level identification is often inaccurate and low abundance pathogens can sometimes be missed. We have developed Taxoner, an open source, taxon assignment pipeline that includes a fast aligner (e.g. Bowtie2) and a comprehensive DNA sequence database. We tested the program on simulated datasets as well as experimental data from Illumina, IonTorrent, and Roche 454 sequencing platforms. We found that Taxoner performs as well as, and often better than BLAST, but requires two orders of magnitude less running time meaning that it can be run on desktop or laptop computers. Taxoner is slower than the approaches that use small marker databases but is more sensitive due the comprehensive reference database. In addition, it can be easily tuned to specific applications using small tailored databases. When applied to metagenomic datasets, Taxoner can provide a functional summary of the genes mapped and can provide strain level identification. Taxoner is written in C for Linux operating systems. The code and documentation are available for research applications at http://code.google.com/p/taxoner.

NutriNet: A Deep Learning Food and Drink Image Recognition System for Dietary Assessment

PubMed Central

Koroušić Seljak, Barbara

2017-01-01

Automatic food image recognition systems are alleviating the process of food-intake estimation and dietary assessment. However, due to the nature of food images, their recognition is a particularly challenging task, which is why traditional approaches in the field have achieved a low classification accuracy. Deep neural networks have outperformed such solutions, and we present a novel approach to the problem of food and drink image detection and recognition that uses a newly-defined deep convolutional neural network architecture, called NutriNet. This architecture was tuned on a recognition dataset containing 225,953 512 × 512 pixel images of 520 different food and drink items from a broad spectrum of food groups, on which we achieved a classification accuracy of 86.72%, along with an accuracy of 94.47% on a detection dataset containing 130,517 images. We also performed a real-world test on a dataset of self-acquired images, combined with images from Parkinson’s disease patients, all taken using a smartphone camera, achieving a top-five accuracy of 55%, which is an encouraging result for real-world images. Additionally, we tested NutriNet on the University of Milano-Bicocca 2016 (UNIMIB2016) food image dataset, on which we improved upon the provided baseline recognition result. An online training component was implemented to continually fine-tune the food and drink recognition model on new images. The model is being used in practice as part of a mobile app for the dietary assessment of Parkinson’s disease patients. PMID:28653995

Selecting AGN through Variability in SN Datasets

NASA Astrophysics Data System (ADS)

Boutsia, K.; Leibundgut, B.; Trevese, D.; Vagnetti, F.

2010-07-01

Variability is a main property of Active Galactic Nuclei (AGN) and it was adopted as a selection criterion using multi epoch surveys conducted for the detection of supernovae (SNe). We have used two SN datasets. First we selected the AXAF field of the STRESS project, centered in the Chandra Deep Field South where, besides the deep X-ray surveys also various optical catalogs exist. Our method yielded 132 variable AGN candidates. We then extended our method including the dataset of the ESSENCE project that has been active for 6 years, producing high quality light curves in the R and I bands. We obtained a sample of ˜4800 variable sources, down to R=22, in the whole 12 deg2 ESSENCE field. Among them, a subsample of ˜500 high priority AGN candidates was created using as secondary criterion the shape of the structure function. In a pilot spectroscopic run we have confirmed the AGN nature for nearly all of our candidates.

DNA Replication Profiling Using Deep Sequencing.

PubMed

Saayman, Xanita; Ramos-Pérez, Cristina; Brown, Grant W

2018-01-01

Profiling of DNA replication during progression through S phase allows a quantitative snap-shot of replication origin usage and DNA replication fork progression. We present a method for using deep sequencing data to profile DNA replication in S. cerevisiae.

Fathead minnow genome sequencing and assembly

EPA Pesticide Factsheets

The dataset provides the URLs for accessing the genome sequence data and two draft assemblies as well as fathead minnow genotyping data associated with estimating the heterozygosity of the in-bred line.This dataset is associated with the following publication:Burns, F., L. Cogburn, G. Ankley , D. Villeneuve , E. Waits , Y. Chang, V. Llaca, S. Deschamps, R. Jackson, and R. Hoke. Sequencing and De novo Draft Assemblies of the Fathead Minnow (Pimphales promelas)Reference Genome. ENVIRONMENTAL TOXICOLOGY AND CHEMISTRY. Society of Environmental Toxicology and Chemistry, Pensacola, FL, USA, 35(1): 212-217, (2016).

Impact of sequencing depth and read length on single cell RNA sequencing data of T cells.

PubMed

Rizzetto, Simone; Eltahla, Auda A; Lin, Peijie; Bull, Rowena; Lloyd, Andrew R; Ho, Joshua W K; Venturi, Vanessa; Luciani, Fabio

2017-10-06

Single cell RNA sequencing (scRNA-seq) provides great potential in measuring the gene expression profiles of heterogeneous cell populations. In immunology, scRNA-seq allowed the characterisation of transcript sequence diversity of functionally relevant T cell subsets, and the identification of the full length T cell receptor (TCRαβ), which defines the specificity against cognate antigens. Several factors, e.g. RNA library capture, cell quality, and sequencing output affect the quality of scRNA-seq data. We studied the effects of read length and sequencing depth on the quality of gene expression profiles, cell type identification, and TCRαβ reconstruction, utilising 1,305 single cells from 8 publically available scRNA-seq datasets, and simulation-based analyses. Gene expression was characterised by an increased number of unique genes identified with short read lengths (<50 bp), but these featured higher technical variability compared to profiles from longer reads. Successful TCRαβ reconstruction was achieved for 6 datasets (81% - 100%) with at least 0.25 millions (PE) reads of length >50 bp, while it failed for datasets with <30 bp reads. Sufficient read length and sequencing depth can control technical noise to enable accurate identification of TCRαβ and gene expression profiles from scRNA-seq data of T cells.

A Deep Learning Framework for Robust and Accurate Prediction of ncRNA-Protein Interactions Using Evolutionary Information.

PubMed

Yi, Hai-Cheng; You, Zhu-Hong; Huang, De-Shuang; Li, Xiao; Jiang, Tong-Hai; Li, Li-Ping

2018-06-01

The interactions between non-coding RNAs (ncRNAs) and proteins play an important role in many biological processes, and their biological functions are primarily achieved by binding with a variety of proteins. High-throughput biological techniques are used to identify protein molecules bound with specific ncRNA, but they are usually expensive and time consuming. Deep learning provides a powerful solution to computationally predict RNA-protein interactions. In this work, we propose the RPI-SAN model by using the deep-learning stacked auto-encoder network to mine the hidden high-level features from RNA and protein sequences and feed them into a random forest (RF) model to predict ncRNA binding proteins. Stacked assembling is further used to improve the accuracy of the proposed method. Four benchmark datasets, including RPI2241, RPI488, RPI1807, and NPInter v2.0, were employed for the unbiased evaluation of five established prediction tools: RPI-Pred, IPMiner, RPISeq-RF, lncPro, and RPI-SAN. The experimental results show that our RPI-SAN model achieves much better performance than other methods, with accuracies of 90.77%, 89.7%, 96.1%, and 99.33%, respectively. It is anticipated that RPI-SAN can be used as an effective computational tool for future biomedical researches and can accurately predict the potential ncRNA-protein interacted pairs, which provides reliable guidance for biological research. Copyright © 2018 The Author(s). Published by Elsevier Inc. All rights reserved.

DeepARG: a deep learning approach for predicting antibiotic resistance genes from metagenomic data.

PubMed

Arango-Argoty, Gustavo; Garner, Emily; Pruden, Amy; Heath, Lenwood S; Vikesland, Peter; Zhang, Liqing

2018-02-01

Growing concerns about increasing rates of antibiotic resistance call for expanded and comprehensive global monitoring. Advancing methods for monitoring of environmental media (e.g., wastewater, agricultural waste, food, and water) is especially needed for identifying potential resources of novel antibiotic resistance genes (ARGs), hot spots for gene exchange, and as pathways for the spread of ARGs and human exposure. Next-generation sequencing now enables direct access and profiling of the total metagenomic DNA pool, where ARGs are typically identified or predicted based on the "best hits" of sequence searches against existing databases. Unfortunately, this approach produces a high rate of false negatives. To address such limitations, we propose here a deep learning approach, taking into account a dissimilarity matrix created using all known categories of ARGs. Two deep learning models, DeepARG-SS and DeepARG-LS, were constructed for short read sequences and full gene length sequences, respectively. Evaluation of the deep learning models over 30 antibiotic resistance categories demonstrates that the DeepARG models can predict ARGs with both high precision (> 0.97) and recall (> 0.90). The models displayed an advantage over the typical best hit approach, yielding consistently lower false negative rates and thus higher overall recall (> 0.9). As more data become available for under-represented ARG categories, the DeepARG models' performance can be expected to be further enhanced due to the nature of the underlying neural networks. Our newly developed ARG database, DeepARG-DB, encompasses ARGs predicted with a high degree of confidence and extensive manual inspection, greatly expanding current ARG repositories. The deep learning models developed here offer more accurate antimicrobial resistance annotation relative to current bioinformatics practice. DeepARG does not require strict cutoffs, which enables identification of a much broader diversity of ARGs. The DeepARG models and database are available as a command line version and as a Web service at http://bench.cs.vt.edu/deeparg .

DEEP MOTIF DASHBOARD: VISUALIZING AND UNDERSTANDING GENOMIC SEQUENCES USING DEEP NEURAL NETWORKS.

PubMed

Lanchantin, Jack; Singh, Ritambhara; Wang, Beilun; Qi, Yanjun

2017-01-01

Deep neural network (DNN) models have recently obtained state-of-the-art prediction accuracy for the transcription factor binding (TFBS) site classification task. However, it remains unclear how these approaches identify meaningful DNA sequence signals and give insights as to why TFs bind to certain locations. In this paper, we propose a toolkit called the Deep Motif Dashboard (DeMo Dashboard) which provides a suite of visualization strategies to extract motifs, or sequence patterns from deep neural network models for TFBS classification. We demonstrate how to visualize and understand three important DNN models: convolutional, recurrent, and convolutional-recurrent networks. Our first visualization method is finding a test sequence's saliency map which uses first-order derivatives to describe the importance of each nucleotide in making the final prediction. Second, considering recurrent models make predictions in a temporal manner (from one end of a TFBS sequence to the other), we introduce temporal output scores, indicating the prediction score of a model over time for a sequential input. Lastly, a class-specific visualization strategy finds the optimal input sequence for a given TFBS positive class via stochastic gradient optimization. Our experimental results indicate that a convolutional-recurrent architecture performs the best among the three architectures. The visualization techniques indicate that CNN-RNN makes predictions by modeling both motifs as well as dependencies among them.

Healthy human gut phageome

PubMed Central

Manrique, Pilar; Bolduc, Benjamin; Walk, Seth T.; van der Oost, John; de Vos, Willem M.; Young, Mark J.

2016-01-01

The role of bacteriophages in influencing the structure and function of the healthy human gut microbiome is unknown. With few exceptions, previous studies have found a high level of heterogeneity in bacteriophages from healthy individuals. To better estimate and identify the shared phageome of humans, we analyzed a deep DNA sequence dataset of active bacteriophages and available metagenomic datasets of the gut bacteriophage community from healthy individuals. We found 23 shared bacteriophages in more than one-half of 64 healthy individuals from around the world. These shared bacteriophages were found in a significantly smaller percentage of individuals with gastrointestinal/irritable bowel disease. A network analysis identified 44 bacteriophage groups of which 9 (20%) were shared in more than one-half of all 64 individuals. These results provide strong evidence of a healthy gut phageome (HGP) in humans. The bacteriophage community in the human gut is a mixture of three classes: a set of core bacteriophages shared among more than one-half of all people, a common set of bacteriophages found in 20–50% of individuals, and a set of bacteriophages that are either rarely shared or unique to a person. We propose that the core and common bacteriophage communities are globally distributed and comprise the HGP, which plays an important role in maintaining gut microbiome structure/function and thereby contributes significantly to human health. PMID:27573828

Healthy human gut phageome.

PubMed

Manrique, Pilar; Bolduc, Benjamin; Walk, Seth T; van der Oost, John; de Vos, Willem M; Young, Mark J

2016-09-13

The role of bacteriophages in influencing the structure and function of the healthy human gut microbiome is unknown. With few exceptions, previous studies have found a high level of heterogeneity in bacteriophages from healthy individuals. To better estimate and identify the shared phageome of humans, we analyzed a deep DNA sequence dataset of active bacteriophages and available metagenomic datasets of the gut bacteriophage community from healthy individuals. We found 23 shared bacteriophages in more than one-half of 64 healthy individuals from around the world. These shared bacteriophages were found in a significantly smaller percentage of individuals with gastrointestinal/irritable bowel disease. A network analysis identified 44 bacteriophage groups of which 9 (20%) were shared in more than one-half of all 64 individuals. These results provide strong evidence of a healthy gut phageome (HGP) in humans. The bacteriophage community in the human gut is a mixture of three classes: a set of core bacteriophages shared among more than one-half of all people, a common set of bacteriophages found in 20-50% of individuals, and a set of bacteriophages that are either rarely shared or unique to a person. We propose that the core and common bacteriophage communities are globally distributed and comprise the HGP, which plays an important role in maintaining gut microbiome structure/function and thereby contributes significantly to human health.

How B-Cell Receptor Repertoire Sequencing Can Be Enriched with Structural Antibody Data

PubMed Central

Kovaltsuk, Aleksandr; Krawczyk, Konrad; Galson, Jacob D.; Kelly, Dominic F.; Deane, Charlotte M.; Trück, Johannes

2017-01-01

Next-generation sequencing of immunoglobulin gene repertoires (Ig-seq) allows the investigation of large-scale antibody dynamics at a sequence level. However, structural information, a crucial descriptor of antibody binding capability, is not collected in Ig-seq protocols. Developing systematic relationships between the antibody sequence information gathered from Ig-seq and low-throughput techniques such as X-ray crystallography could radically improve our understanding of antibodies. The mapping of Ig-seq datasets to known antibody structures can indicate structurally, and perhaps functionally, uncharted areas. Furthermore, contrasting naïve and antigenically challenged datasets using structural antibody descriptors should provide insights into antibody maturation. As the number of antibody structures steadily increases and more and more Ig-seq datasets become available, the opportunities that arise from combining the two types of information increase as well. Here, we review how these data types enrich one another and show potential for advancing our knowledge of the immune system and improving antibody engineering. PMID:29276518

Treetrimmer: a method for phylogenetic dataset size reduction.

PubMed

Maruyama, Shinichiro; Eveleigh, Robert J M; Archibald, John M

2013-04-12

With rapid advances in genome sequencing and bioinformatics, it is now possible to generate phylogenetic trees containing thousands of operational taxonomic units (OTUs) from a wide range of organisms. However, use of rigorous tree-building methods on such large datasets is prohibitive and manual 'pruning' of sequence alignments is time consuming and raises concerns over reproducibility. There is a need for bioinformatic tools with which to objectively carry out such pruning procedures. Here we present 'TreeTrimmer', a bioinformatics procedure that removes unnecessary redundancy in large phylogenetic datasets, alleviating the size effect on more rigorous downstream analyses. The method identifies and removes user-defined 'redundant' sequences, e.g., orthologous sequences from closely related organisms and 'recently' evolved lineage-specific paralogs. Representative OTUs are retained for more rigorous re-analysis. TreeTrimmer reduces the OTU density of phylogenetic trees without sacrificing taxonomic diversity while retaining the original tree topology, thereby speeding up downstream computer-intensive analyses, e.g., Bayesian and maximum likelihood tree reconstructions, in a reproducible fashion.

Relationships between palaeogeography and opal occurrence in Australia: A data-mining approach

NASA Astrophysics Data System (ADS)

Landgrebe, T. C. W.; Merdith, A.; Dutkiewicz, A.; Müller, R. D.

2013-07-01

Age-coded multi-layered geological datasets are becoming increasingly prevalent with the surge in open-access geodata, yet there are few methodologies for extracting geological information and knowledge from these data. We present a novel methodology, based on the open-source GPlates software in which age-coded digital palaeogeographic maps are used to “data-mine” spatio-temporal patterns related to the occurrence of Australian opal. Our aim is to test the concept that only a particular sequence of depositional/erosional environments may lead to conditions suitable for the formation of gem quality sedimentary opal. Time-varying geographic environment properties are extracted from a digital palaeogeographic dataset of the eastern Australian Great Artesian Basin (GAB) at 1036 opal localities. We obtain a total of 52 independent ordinal sequences sampling 19 time slices from the Early Cretaceous to the present-day. We find that 95% of the known opal deposits are tied to only 27 sequences all comprising fluvial and shallow marine depositional sequences followed by a prolonged phase of erosion. We then map the total area of the GAB that matches these 27 opal-specific sequences, resulting in an opal-prospective region of only about 10% of the total area of the basin. The key patterns underlying this association involve only a small number of key environmental transitions. We demonstrate that these key associations are generally absent at arbitrary locations in the basin. This new methodology allows for the simplification of a complex time-varying geological dataset into a single map view, enabling straightforward application for opal exploration and for future co-assessment with other datasets/geological criteria. This approach may help unravel the poorly understood opal formation process using an empirical spatio-temporal data-mining methodology and readily available datasets to aid hypothesis testing.

Metabarcoding of marine nematodes – evaluation of reference datasets used in tree-based taxonomy assignment approach

PubMed Central

2016-01-01

Abstract Background Metabarcoding is becoming a common tool used to assess and compare diversity of organisms in environmental samples. Identification of OTUs is one of the critical steps in the process and several taxonomy assignment methods were proposed to accomplish this task. This publication evaluates the quality of reference datasets, alongside with several alignment and phylogeny inference methods used in one of the taxonomy assignment methods, called tree-based approach. This approach assigns anonymous OTUs to taxonomic categories based on relative placements of OTUs and reference sequences on the cladogram and support that these placements receive. New information In tree-based taxonomy assignment approach, reliable identification of anonymous OTUs is based on their placement in monophyletic and highly supported clades together with identified reference taxa. Therefore, it requires high quality reference dataset to be used. Resolution of phylogenetic trees is strongly affected by the presence of erroneous sequences as well as alignment and phylogeny inference methods used in the process. Two preparation steps are essential for the successful application of tree-based taxonomy assignment approach. Curated collections of genetic information do include erroneous sequences. These sequences have detrimental effect on the resolution of cladograms used in tree-based approach. They must be identified and excluded from the reference dataset beforehand. Various combinations of multiple sequence alignment and phylogeny inference methods provide cladograms with different topology and bootstrap support. These combinations of methods need to be tested in order to determine the one that gives highest resolution for the particular reference dataset. Completing the above mentioned preparation steps is expected to decrease the number of unassigned OTUs and thus improve the results of the tree-based taxonomy assignment approach. PMID:27932919

Metabarcoding of marine nematodes - evaluation of reference datasets used in tree-based taxonomy assignment approach.

PubMed

Holovachov, Oleksandr

2016-01-01

Metabarcoding is becoming a common tool used to assess and compare diversity of organisms in environmental samples. Identification of OTUs is one of the critical steps in the process and several taxonomy assignment methods were proposed to accomplish this task. This publication evaluates the quality of reference datasets, alongside with several alignment and phylogeny inference methods used in one of the taxonomy assignment methods, called tree-based approach. This approach assigns anonymous OTUs to taxonomic categories based on relative placements of OTUs and reference sequences on the cladogram and support that these placements receive. In tree-based taxonomy assignment approach, reliable identification of anonymous OTUs is based on their placement in monophyletic and highly supported clades together with identified reference taxa. Therefore, it requires high quality reference dataset to be used. Resolution of phylogenetic trees is strongly affected by the presence of erroneous sequences as well as alignment and phylogeny inference methods used in the process. Two preparation steps are essential for the successful application of tree-based taxonomy assignment approach. Curated collections of genetic information do include erroneous sequences. These sequences have detrimental effect on the resolution of cladograms used in tree-based approach. They must be identified and excluded from the reference dataset beforehand.Various combinations of multiple sequence alignment and phylogeny inference methods provide cladograms with different topology and bootstrap support. These combinations of methods need to be tested in order to determine the one that gives highest resolution for the particular reference dataset.Completing the above mentioned preparation steps is expected to decrease the number of unassigned OTUs and thus improve the results of the tree-based taxonomy assignment approach.

TranslatomeDB: a comprehensive database and cloud-based analysis platform for translatome sequencing data.

PubMed

Liu, Wanting; Xiang, Lunping; Zheng, Tingkai; Jin, Jingjie; Zhang, Gong

2018-01-04

Translation is a key regulatory step, linking transcriptome and proteome. Two major methods of translatome investigations are RNC-seq (sequencing of translating mRNA) and Ribo-seq (ribosome profiling). To facilitate the investigation of translation, we built a comprehensive database TranslatomeDB (http://www.translatomedb.net/) which provides collection and integrated analysis of published and user-generated translatome sequencing data. The current version includes 2453 Ribo-seq, 10 RNC-seq and their 1394 corresponding mRNA-seq datasets in 13 species. The database emphasizes the analysis functions in addition to the dataset collections. Differential gene expression (DGE) analysis can be performed between any two datasets of same species and type, both on transcriptome and translatome levels. The translation indices translation ratios, elongation velocity index and translational efficiency can be calculated to quantitatively evaluate translational initiation efficiency and elongation velocity, respectively. All datasets were analyzed using a unified, robust, accurate and experimentally-verifiable pipeline based on the FANSe3 mapping algorithm and edgeR for DGE analyzes. TranslatomeDB also allows users to upload their own datasets and utilize the identical unified pipeline to analyze their data. We believe that our TranslatomeDB is a comprehensive platform and knowledgebase on translatome and proteome research, releasing the biologists from complex searching, analyzing and comparing huge sequencing data without needing local computational power. © The Author(s) 2017. Published by Oxford University Press on behalf of Nucleic Acids Research.

Analysis of plant-derived miRNAs in animal small RNA datasets

PubMed Central

2012-01-01

Background Plants contain significant quantities of small RNAs (sRNAs) derived from various sRNA biogenesis pathways. Many of these sRNAs play regulatory roles in plants. Previous analysis revealed that numerous sRNAs in corn, rice and soybean seeds have high sequence similarity to animal genes. However, exogenous RNA is considered to be unstable within the gastrointestinal tract of many animals, thus limiting potential for any adverse effects from consumption of dietary RNA. A recent paper reported that putative plant miRNAs were detected in animal plasma and serum, presumably acquired through ingestion, and may have a functional impact in the consuming organisms. Results To address the question of how common this phenomenon could be, we searched for plant miRNAs sequences in public sRNA datasets from various tissues of mammals, chicken and insects. Our analyses revealed that plant miRNAs were present in the animal sRNA datasets, and significantly miR168 was extremely over-represented. Furthermore, all or nearly all (>96%) miR168 sequences were monocot derived for most datasets, including datasets for two insects reared on dicot plants in their respective experiments. To investigate if plant-derived miRNAs, including miR168, could accumulate and move systemically in insects, we conducted insect feeding studies for three insects including corn rootworm, which has been shown to be responsive to plant-produced long double-stranded RNAs. Conclusions Our analyses suggest that the observed plant miRNAs in animal sRNA datasets can originate in the process of sequencing, and that accumulation of plant miRNAs via dietary exposure is not universal in animals. PMID:22873950

An Alignment-Free Algorithm in Comparing the Similarity of Protein Sequences Based on Pseudo-Markov Transition Probabilities among Amino Acids

PubMed Central

Li, Yushuang; Yang, Jiasheng; Zhang, Yi

2016-01-01

In this paper, we have proposed a novel alignment-free method for comparing the similarity of protein sequences. We first encode a protein sequence into a 440 dimensional feature vector consisting of a 400 dimensional Pseudo-Markov transition probability vector among the 20 amino acids, a 20 dimensional content ratio vector, and a 20 dimensional position ratio vector of the amino acids in the sequence. By evaluating the Euclidean distances among the representing vectors, we compare the similarity of protein sequences. We then apply this method into the ND5 dataset consisting of the ND5 protein sequences of 9 species, and the F10 and G11 datasets representing two of the xylanases containing glycoside hydrolase families, i.e., families 10 and 11. As a result, our method achieves a correlation coefficient of 0.962 with the canonical protein sequence aligner ClustalW in the ND5 dataset, much higher than those of other 5 popular alignment-free methods. In addition, we successfully separate the xylanases sequences in the F10 family and the G11 family and illustrate that the F10 family is more heat stable than the G11 family, consistent with a few previous studies. Moreover, we prove mathematically an identity equation involving the Pseudo-Markov transition probability vector and the amino acids content ratio vector. PMID:27918587

The Transcriptome Analysis and Comparison Explorer--T-ACE: a platform-independent, graphical tool to process large RNAseq datasets of non-model organisms.

PubMed

Philipp, E E R; Kraemer, L; Mountfort, D; Schilhabel, M; Schreiber, S; Rosenstiel, P

2012-03-15

Next generation sequencing (NGS) technologies allow a rapid and cost-effective compilation of large RNA sequence datasets in model and non-model organisms. However, the storage and analysis of transcriptome information from different NGS platforms is still a significant bottleneck, leading to a delay in data dissemination and subsequent biological understanding. Especially database interfaces with transcriptome analysis modules going beyond mere read counts are missing. Here, we present the Transcriptome Analysis and Comparison Explorer (T-ACE), a tool designed for the organization and analysis of large sequence datasets, and especially suited for transcriptome projects of non-model organisms with little or no a priori sequence information. T-ACE offers a TCL-based interface, which accesses a PostgreSQL database via a php-script. Within T-ACE, information belonging to single sequences or contigs, such as annotation or read coverage, is linked to the respective sequence and immediately accessible. Sequences and assigned information can be searched via keyword- or BLAST-search. Additionally, T-ACE provides within and between transcriptome analysis modules on the level of expression, GO terms, KEGG pathways and protein domains. Results are visualized and can be easily exported for external analysis. We developed T-ACE for laboratory environments, which have only a limited amount of bioinformatics support, and for collaborative projects in which different partners work on the same dataset from different locations or platforms (Windows/Linux/MacOS). For laboratories with some experience in bioinformatics and programming, the low complexity of the database structure and open-source code provides a framework that can be customized according to the different needs of the user and transcriptome project.

Benchmark datasets for phylogenomic pipeline validation, applications for foodborne pathogen surveillance

PubMed Central

Rand, Hugh; Shumway, Martin; Trees, Eija K.; Simmons, Mustafa; Agarwala, Richa; Davis, Steven; Tillman, Glenn E.; Defibaugh-Chavez, Stephanie; Carleton, Heather A.; Klimke, William A.; Katz, Lee S.

2017-01-01

Background As next generation sequence technology has advanced, there have been parallel advances in genome-scale analysis programs for determining evolutionary relationships as proxies for epidemiological relationship in public health. Most new programs skip traditional steps of ortholog determination and multi-gene alignment, instead identifying variants across a set of genomes, then summarizing results in a matrix of single-nucleotide polymorphisms or alleles for standard phylogenetic analysis. However, public health authorities need to document the performance of these methods with appropriate and comprehensive datasets so they can be validated for specific purposes, e.g., outbreak surveillance. Here we propose a set of benchmark datasets to be used for comparison and validation of phylogenomic pipelines. Methods We identified four well-documented foodborne pathogen events in which the epidemiology was concordant with routine phylogenomic analyses (reference-based SNP and wgMLST approaches). These are ideal benchmark datasets, as the trees, WGS data, and epidemiological data for each are all in agreement. We have placed these sequence data, sample metadata, and “known” phylogenetic trees in publicly-accessible databases and developed a standard descriptive spreadsheet format describing each dataset. To facilitate easy downloading of these benchmarks, we developed an automated script that uses the standard descriptive spreadsheet format. Results Our “outbreak” benchmark datasets represent the four major foodborne bacterial pathogens (Listeria monocytogenes, Salmonella enterica, Escherichia coli, and Campylobacter jejuni) and one simulated dataset where the “known tree” can be accurately called the “true tree”. The downloading script and associated table files are available on GitHub: https://github.com/WGS-standards-and-analysis/datasets. Discussion These five benchmark datasets will help standardize comparison of current and future phylogenomic pipelines, and facilitate important cross-institutional collaborations. Our work is part of a global effort to provide collaborative infrastructure for sequence data and analytic tools—we welcome additional benchmark datasets in our recommended format, and, if relevant, we will add these on our GitHub site. Together, these datasets, dataset format, and the underlying GitHub infrastructure present a recommended path for worldwide standardization of phylogenomic pipelines. PMID:29372115

Benchmark datasets for phylogenomic pipeline validation, applications for foodborne pathogen surveillance.

PubMed

Timme, Ruth E; Rand, Hugh; Shumway, Martin; Trees, Eija K; Simmons, Mustafa; Agarwala, Richa; Davis, Steven; Tillman, Glenn E; Defibaugh-Chavez, Stephanie; Carleton, Heather A; Klimke, William A; Katz, Lee S

2017-01-01

As next generation sequence technology has advanced, there have been parallel advances in genome-scale analysis programs for determining evolutionary relationships as proxies for epidemiological relationship in public health. Most new programs skip traditional steps of ortholog determination and multi-gene alignment, instead identifying variants across a set of genomes, then summarizing results in a matrix of single-nucleotide polymorphisms or alleles for standard phylogenetic analysis. However, public health authorities need to document the performance of these methods with appropriate and comprehensive datasets so they can be validated for specific purposes, e.g., outbreak surveillance. Here we propose a set of benchmark datasets to be used for comparison and validation of phylogenomic pipelines. We identified four well-documented foodborne pathogen events in which the epidemiology was concordant with routine phylogenomic analyses (reference-based SNP and wgMLST approaches). These are ideal benchmark datasets, as the trees, WGS data, and epidemiological data for each are all in agreement. We have placed these sequence data, sample metadata, and "known" phylogenetic trees in publicly-accessible databases and developed a standard descriptive spreadsheet format describing each dataset. To facilitate easy downloading of these benchmarks, we developed an automated script that uses the standard descriptive spreadsheet format. Our "outbreak" benchmark datasets represent the four major foodborne bacterial pathogens ( Listeria monocytogenes , Salmonella enterica , Escherichia coli , and Campylobacter jejuni ) and one simulated dataset where the "known tree" can be accurately called the "true tree". The downloading script and associated table files are available on GitHub: https://github.com/WGS-standards-and-analysis/datasets. These five benchmark datasets will help standardize comparison of current and future phylogenomic pipelines, and facilitate important cross-institutional collaborations. Our work is part of a global effort to provide collaborative infrastructure for sequence data and analytic tools-we welcome additional benchmark datasets in our recommended format, and, if relevant, we will add these on our GitHub site. Together, these datasets, dataset format, and the underlying GitHub infrastructure present a recommended path for worldwide standardization of phylogenomic pipelines.

Transcriptome-wide profiling and expression analysis of transcription factor families in a liverwort, Marchantia polymorpha

PubMed Central

2013-01-01

Background Transcription factors (TFs) are vital elements that regulate transcription and the spatio-temporal expression of genes, thereby ensuring the accurate development and functioning of an organism. The identification of TF-encoding genes in a liverwort, Marchantia polymorpha, offers insights into TF organization in the members of the most basal lineages of land plants (embryophytes). Therefore, a comparison of Marchantia TF genes with other land plants (monocots, dicots, bryophytes) and algae (chlorophytes, rhodophytes) provides the most comprehensive view of the rates of expansion or contraction of TF genes in plant evolution. Results In this study, we report the identification of TF-encoding transcripts in M. polymorpha for the first time, as evidenced by deep RNA sequencing data. In total, 3,471 putative TF encoding transcripts, distributed in 80 families, were identified, representing 7.4% of the generated Marchantia gametophytic transcriptome dataset. Overall, TF basic functions and distribution across families appear to be conserved when compared to other plant species. However, it is of interest to observe the genesis of novel sequences in 24 TF families and the apparent termination of 2 TF families with the emergence of Marchantia. Out of 24 TF families, 6 are known to be associated with plant reproductive development processes. We also examined the expression pattern of these TF-encoding transcripts in six male and female developmental stages in vegetative and reproductive gametophytic tissues of Marchantia. Conclusions The analysis highlighted the importance of Marchantia, a model plant system, in an evolutionary context. The dataset generated here provides a scientific resource for TF gene discovery and other comparative evolutionary studies of land plants. PMID:24365221

Xenopus in Space and Time: Fossils, Node Calibrations, Tip-Dating, and Paleobiogeography.

PubMed

Cannatella, David

2015-01-01

Published data from DNA sequences, morphology of 11 extant and 15 extinct frog taxa, and stratigraphic ranges of fossils were integrated to open a window into the deep-time evolution of Xenopus. The ages and morphological characters of fossils were used as independent datasets to calibrate a chronogram. We found that DNA sequences, either alone or in combination with morphological data and fossils, tended to support a close relationship between Xenopus and Hymenochirus, although in some analyses this topology was not significantly better than the Pipa + Hymenochirus topology. Analyses that excluded DNA data found strong support for the Pipa + Hymenochirus tree. The criterion for selecting the maximum age of the calibration prior influenced the age estimates, and our age estimates of early divergences in the tree of frogs are substantially younger than those of published studies. Node-dating and tip-dating calibrations, either alone or in combination, yielded older dates for nodes than did a root calibration alone. Our estimates of divergence times indicate that overwater dispersal, rather than vicariance due to the splitting of Africa and South America, may explain the presence of Xenopus in Africa and its closest fossil relatives in South America.

Construction and Analysis of Functional Networks in the Gut Microbiome of Type 2 Diabetes Patients.

PubMed

Li, Lianshuo; Wang, Zicheng; He, Peng; Ma, Shining; Du, Jie; Jiang, Rui

2016-10-01

Although networks of microbial species have been widely used in the analysis of 16S rRNA sequencing data of a microbiome, the construction and analysis of a complete microbial gene network are in general problematic because of the large number of microbial genes in metagenomics studies. To overcome this limitation, we propose to map microbial genes to functional units, including KEGG orthologous groups and the evolutionary genealogy of genes: Non-supervised Orthologous Groups (eggNOG) orthologous groups, to enable the construction and analysis of a microbial functional network. We devised two statistical methods to infer pairwise relationships between microbial functional units based on a deep sequencing dataset of gut microbiome from type 2 diabetes (T2D) patients as well as healthy controls. Networks containing such functional units and their significant interactions were constructed subsequently. We conducted a variety of analyses of global properties, local properties, and functional modules in the resulting functional networks. Our data indicate that besides the observations consistent with the current knowledge, this study provides novel biological insights into the gut microbiome associated with T2D. Copyright © 2016. Production and hosting by Elsevier Ltd.

Residual Shuffling Convolutional Neural Networks for Deep Semantic Image Segmentation Using Multi-Modal Data

NASA Astrophysics Data System (ADS)

Chen, K.; Weinmann, M.; Gao, X.; Yan, M.; Hinz, S.; Jutzi, B.; Weinmann, M.

2018-05-01

In this paper, we address the deep semantic segmentation of aerial imagery based on multi-modal data. Given multi-modal data composed of true orthophotos and the corresponding Digital Surface Models (DSMs), we extract a variety of hand-crafted radiometric and geometric features which are provided separately and in different combinations as input to a modern deep learning framework. The latter is represented by a Residual Shuffling Convolutional Neural Network (RSCNN) combining the characteristics of a Residual Network with the advantages of atrous convolution and a shuffling operator to achieve a dense semantic labeling. Via performance evaluation on a benchmark dataset, we analyze the value of different feature sets for the semantic segmentation task. The derived results reveal that the use of radiometric features yields better classification results than the use of geometric features for the considered dataset. Furthermore, the consideration of data on both modalities leads to an improvement of the classification results. However, the derived results also indicate that the use of all defined features is less favorable than the use of selected features. Consequently, data representations derived via feature extraction and feature selection techniques still provide a gain if used as the basis for deep semantic segmentation.

Intervertebral disc detection in X-ray images using faster R-CNN.

PubMed

Ruhan Sa; Owens, William; Wiegand, Raymond; Studin, Mark; Capoferri, Donald; Barooha, Kenneth; Greaux, Alexander; Rattray, Robert; Hutton, Adam; Cintineo, John; Chaudhary, Vipin

2017-07-01

Automatic identification of specific osseous landmarks on the spinal radiograph can be used to automate calculations for correcting ligament instability and injury, which affect 75% of patients injured in motor vehicle accidents. In this work, we propose to use deep learning based object detection method as the first step towards identifying landmark points in lateral lumbar X-ray images. The significant breakthrough of deep learning technology has made it a prevailing choice for perception based applications, however, the lack of large annotated training dataset has brought challenges to utilizing the technology in medical image processing field. In this work, we propose to fine tune a deep network, Faster-RCNN, a state-of-the-art deep detection network in natural image domain, using small annotated clinical datasets. In the experiment we show that, by using only 81 lateral lumbar X-Ray training images, one can achieve much better performance compared to traditional sliding window detection method on hand crafted features. Furthermore, we fine-tuned the network using 974 training images and tested on 108 images, which achieved average precision of 0.905 with average computation time of 3 second per image, which greatly outperformed traditional methods in terms of accuracy and efficiency.

Adding the missing piece: Spitzer imaging of the HSC-Deep/PFS fields

NASA Astrophysics Data System (ADS)

Sajina, Anna; Bezanson, Rachel; Capak, Peter; Egami, Eiichi; Fan, Xiaohui; Farrah, Duncan; Greene, Jenny; Goulding, Andy; Lacy, Mark; Lin, Yen-Ting; Liu, Xin; Marchesini, Danilo; Moutard, Thibaud; Ono, Yoshiaki; Ouchi, Masami; Sawicki, Marcin; Strauss, Michael; Surace, Jason; Whitaker, Katherine

2018-05-01

We propose to observe a total of 7sq.deg. to complete the Spitzer-IRAC coverage of the HSC-Deep survey fields. These fields are the sites of the PrimeFocusSpectrograph (PFS) galaxy evolution survey which will provide spectra of wide wavelength range and resolution for almost all M* galaxies at z 0.7-1.7, and extend out to z 7 for targeted samples. Our fields already have deep broadband and narrowband photometry in 12 bands spanning from u through K and a wealth of other ancillary data. We propose completing the matching depth IRAC observations in the extended COSMOS, ELAIS-N1 and Deep2-3 fields. By complementing existing Spitzer coverage, this program will lead to an unprecedended in spectro-photometric coverage dataset across a total of 15 sq.deg. This dataset will have significant legacy value as it samples a large enough cosmic volume to be representative of the full range of environments, but also doing so with sufficient information content per galaxy to confidently derive stellar population characteristics. This enables detailed studies of the growth and quenching of galaxies and their supermassive black holes in the context of a galaxy's local and large scale environment.

A Deep Learning based Approach to Reduced Order Modeling of Fluids using LSTM Neural Networks

NASA Astrophysics Data System (ADS)

Mohan, Arvind; Gaitonde, Datta

2017-11-01

Reduced Order Modeling (ROM) can be used as surrogates to prohibitively expensive simulations to model flow behavior for long time periods. ROM is predicated on extracting dominant spatio-temporal features of the flow from CFD or experimental datasets. We explore ROM development with a deep learning approach, which comprises of learning functional relationships between different variables in large datasets for predictive modeling. Although deep learning and related artificial intelligence based predictive modeling techniques have shown varied success in other fields, such approaches are in their initial stages of application to fluid dynamics. Here, we explore the application of the Long Short Term Memory (LSTM) neural network to sequential data, specifically to predict the time coefficients of Proper Orthogonal Decomposition (POD) modes of the flow for future timesteps, by training it on data at previous timesteps. The approach is demonstrated by constructing ROMs of several canonical flows. Additionally, we show that statistical estimates of stationarity in the training data can indicate a priori how amenable a given flow-field is to this approach. Finally, the potential and limitations of deep learning based ROM approaches will be elucidated and further developments discussed.

Deep machine learning provides state-of-the-art performance in image-based plant phenotyping.

PubMed

Pound, Michael P; Atkinson, Jonathan A; Townsend, Alexandra J; Wilson, Michael H; Griffiths, Marcus; Jackson, Aaron S; Bulat, Adrian; Tzimiropoulos, Georgios; Wells, Darren M; Murchie, Erik H; Pridmore, Tony P; French, Andrew P

2017-10-01

In plant phenotyping, it has become important to be able to measure many features on large image sets in order to aid genetic discovery. The size of the datasets, now often captured robotically, often precludes manual inspection, hence the motivation for finding a fully automated approach. Deep learning is an emerging field that promises unparalleled results on many data analysis problems. Building on artificial neural networks, deep approaches have many more hidden layers in the network, and hence have greater discriminative and predictive power. We demonstrate the use of such approaches as part of a plant phenotyping pipeline. We show the success offered by such techniques when applied to the challenging problem of image-based plant phenotyping and demonstrate state-of-the-art results (>97% accuracy) for root and shoot feature identification and localization. We use fully automated trait identification using deep learning to identify quantitative trait loci in root architecture datasets. The majority (12 out of 14) of manually identified quantitative trait loci were also discovered using our automated approach based on deep learning detection to locate plant features. We have shown deep learning-based phenotyping to have very good detection and localization accuracy in validation and testing image sets. We have shown that such features can be used to derive meaningful biological traits, which in turn can be used in quantitative trait loci discovery pipelines. This process can be completely automated. We predict a paradigm shift in image-based phenotyping bought about by such deep learning approaches, given sufficient training sets. © The Authors 2017. Published by Oxford University Press.

A deep convolutional neural network approach to single-particle recognition in cryo-electron microscopy.

PubMed

Zhu, Yanan; Ouyang, Qi; Mao, Youdong

2017-07-21

Single-particle cryo-electron microscopy (cryo-EM) has become a mainstream tool for the structural determination of biological macromolecular complexes. However, high-resolution cryo-EM reconstruction often requires hundreds of thousands of single-particle images. Particle extraction from experimental micrographs thus can be laborious and presents a major practical bottleneck in cryo-EM structural determination. Existing computational methods for particle picking often use low-resolution templates for particle matching, making them susceptible to reference-dependent bias. It is critical to develop a highly efficient template-free method for the automatic recognition of particle images from cryo-EM micrographs. We developed a deep learning-based algorithmic framework, DeepEM, for single-particle recognition from noisy cryo-EM micrographs, enabling automated particle picking, selection and verification in an integrated fashion. The kernel of DeepEM is built upon a convolutional neural network (CNN) composed of eight layers, which can be recursively trained to be highly "knowledgeable". Our approach exhibits an improved performance and accuracy when tested on the standard KLH dataset. Application of DeepEM to several challenging experimental cryo-EM datasets demonstrated its ability to avoid the selection of un-wanted particles and non-particles even when true particles contain fewer features. The DeepEM methodology, derived from a deep CNN, allows automated particle extraction from raw cryo-EM micrographs in the absence of a template. It demonstrates an improved performance, objectivity and accuracy. Application of this novel method is expected to free the labor involved in single-particle verification, significantly improving the efficiency of cryo-EM data processing.

Enhanced sensitivity for detection of low-level germline mosaic RB1 mutations in sporadic retinoblastoma cases using deep semiconductor sequencing.

PubMed

Chen, Zhao; Moran, Kimberly; Richards-Yutz, Jennifer; Toorens, Erik; Gerhart, Daniel; Ganguly, Tapan; Shields, Carol L; Ganguly, Arupa

2014-03-01

Sporadic retinoblastoma (RB) is caused by de novo mutations in the RB1 gene. Often, these mutations are present as mosaic mutations that cannot be detected by Sanger sequencing. Next-generation deep sequencing allows unambiguous detection of the mosaic mutations in lymphocyte DNA. Deep sequencing of the RB1 gene on lymphocyte DNA from 20 bilateral and 70 unilateral RB cases was performed, where Sanger sequencing excluded the presence of mutations. The individual exons of the RB1 gene from each sample were amplified, pooled, ligated to barcoded adapters, and sequenced using semiconductor sequencing on an Ion Torrent Personal Genome Machine. Six low-level mosaic mutations were identified in bilateral RB and four in unilateral RB cases. The incidence of low-level mosaic mutation was estimated to be 30% and 6%, respectively, in sporadic bilateral and unilateral RB cases, previously classified as mutation negative. The frequency of point mutations detectable in lymphocyte DNA increased from 96% to 97% for bilateral RB and from 13% to 18% for unilateral RB. The use of deep sequencing technology increased the sensitivity of the detection of low-level germline mosaic mutations in the RB1 gene. This finding has significant implications for improved clinical diagnosis, genetic counseling, surveillance, and management of RB. © 2013 WILEY PERIODICALS, INC.

Deep Learning to Predict Falls in Older Adults Based on Daily-Life Trunk Accelerometry.

PubMed

Nait Aicha, Ahmed; Englebienne, Gwenn; van Schooten, Kimberley S; Pijnappels, Mirjam; Kröse, Ben

2018-05-22

Early detection of high fall risk is an essential component of fall prevention in older adults. Wearable sensors can provide valuable insight into daily-life activities; biomechanical features extracted from such inertial data have been shown to be of added value for the assessment of fall risk. Body-worn sensors such as accelerometers can provide valuable insight into fall risk. Currently, biomechanical features derived from accelerometer data are used for the assessment of fall risk. Here, we studied whether deep learning methods from machine learning are suited to automatically derive features from raw accelerometer data that assess fall risk. We used an existing dataset of 296 older adults. We compared the performance of three deep learning model architectures (convolutional neural network (CNN), long short-term memory (LSTM) and a combination of these two (ConvLSTM)) to each other and to a baseline model with biomechanical features on the same dataset. The results show that the deep learning models in a single-task learning mode are strong in recognition of identity of the subject, but that these models only slightly outperform the baseline method on fall risk assessment. When using multi-task learning, with gender and age as auxiliary tasks, deep learning models perform better. We also found that preprocessing of the data resulted in the best performance (AUC = 0.75). We conclude that deep learning models, and in particular multi-task learning, effectively assess fall risk on the basis of wearable sensor data.

Deep Learning to Predict Falls in Older Adults Based on Daily-Life Trunk Accelerometry

PubMed Central

Englebienne, Gwenn; Pijnappels, Mirjam

2018-01-01

Early detection of high fall risk is an essential component of fall prevention in older adults. Wearable sensors can provide valuable insight into daily-life activities; biomechanical features extracted from such inertial data have been shown to be of added value for the assessment of fall risk. Body-worn sensors such as accelerometers can provide valuable insight into fall risk. Currently, biomechanical features derived from accelerometer data are used for the assessment of fall risk. Here, we studied whether deep learning methods from machine learning are suited to automatically derive features from raw accelerometer data that assess fall risk. We used an existing dataset of 296 older adults. We compared the performance of three deep learning model architectures (convolutional neural network (CNN), long short-term memory (LSTM) and a combination of these two (ConvLSTM)) to each other and to a baseline model with biomechanical features on the same dataset. The results show that the deep learning models in a single-task learning mode are strong in recognition of identity of the subject, but that these models only slightly outperform the baseline method on fall risk assessment. When using multi-task learning, with gender and age as auxiliary tasks, deep learning models perform better. We also found that preprocessing of the data resulted in the best performance (AUC = 0.75). We conclude that deep learning models, and in particular multi-task learning, effectively assess fall risk on the basis of wearable sensor data. PMID:29786659

cGRNB: a web server for building combinatorial gene regulatory networks through integrated engineering of seed-matching sequence information and gene expression datasets.

PubMed

Xu, Huayong; Yu, Hui; Tu, Kang; Shi, Qianqian; Wei, Chaochun; Li, Yuan-Yuan; Li, Yi-Xue

2013-01-01

We are witnessing rapid progress in the development of methodologies for building the combinatorial gene regulatory networks involving both TFs (Transcription Factors) and miRNAs (microRNAs). There are a few tools available to do these jobs but most of them are not easy to use and not accessible online. A web server is especially needed in order to allow users to upload experimental expression datasets and build combinatorial regulatory networks corresponding to their particular contexts. In this work, we compiled putative TF-gene, miRNA-gene and TF-miRNA regulatory relationships from forward-engineering pipelines and curated them as built-in data libraries. We streamlined the R codes of our two separate forward-and-reverse engineering algorithms for combinatorial gene regulatory network construction and formalized them as two major functional modules. As a result, we released the cGRNB (combinatorial Gene Regulatory Networks Builder): a web server for constructing combinatorial gene regulatory networks through integrated engineering of seed-matching sequence information and gene expression datasets. The cGRNB enables two major network-building modules, one for MPGE (miRNA-perturbed gene expression) datasets and the other for parallel miRNA/mRNA expression datasets. A miRNA-centered two-layer combinatorial regulatory cascade is the output of the first module and a comprehensive genome-wide network involving all three types of combinatorial regulations (TF-gene, TF-miRNA, and miRNA-gene) are the output of the second module. In this article we propose cGRNB, a web server for building combinatorial gene regulatory networks through integrated engineering of seed-matching sequence information and gene expression datasets. Since parallel miRNA/mRNA expression datasets are rapidly accumulated by the advance of next-generation sequencing techniques, cGRNB will be very useful tool for researchers to build combinatorial gene regulatory networks based on expression datasets. The cGRNB web-server is free and available online at http://www.scbit.org/cgrnb.

Draft Genome Sequence of Deep-Sea Alteromonas sp. Strain V450 Isolated from the Marine Sponge Leiodermatium sp.

PubMed Central

Barrett, Nolan H.; McCarthy, Peter J.

2017-01-01

ABSTRACT The proteobacterium Alteromonas sp. strain V450 was isolated from the Atlantic deep-sea sponge Leiodermatium sp. Here, we report the draft genome sequence of this strain, with a genome size of approx. 4.39 Mb and a G+C content of 44.01%. The results will aid deep-sea microbial ecology, evolution, and sponge-microbe association studies. PMID:28153886

Discovery of Culex pipiens associated tunisia virus: a new ssRNA(+) virus representing a new insect associated virus family

PubMed Central

Bigot, Diane; Atyame, Célestine M; Weill, Mylène; Justy, Fabienne

2018-01-01

Abstract In the global context of arboviral emergence, deep sequencing unlocks the discovery of new mosquito-borne viruses. Mosquitoes of the species Culex pipiens, C. torrentium, and C. hortensis were sampled from 22 locations worldwide for transcriptomic analyses. A virus discovery pipeline was used to analyze the dataset of 0.7 billion reads comprising 22 individual transcriptomes. Two closely related 6.8 kb viral genomes were identified in C. pipiens and named as Culex pipiens associated tunisia virus (CpATV) strains Ayed and Jedaida. The CpATV genome contained four ORFs. ORF1 possessed helicase and RNA-dependent RNA polymerase (RdRp) domains related to new viral sequences recently found mainly in dipterans. ORF2 and 4 contained a capsid protein domain showing strong homology with Virgaviridae plant viruses. ORF3 displayed similarities with eukaryotic Rhoptry domain and a merozoite surface protein (MSP7) domain only found in mosquito-transmitted Plasmodium, suggesting possible interactions between CpATV and vertebrate cells. Estimation of a strong purifying selection exerted on each ORFs and the presence of a polymorphism maintained in the coding region of ORF3 suggested that both CpATV sequences are genuine functional viruses. CpATV is part of an entirely new and highly diversified group of viruses recently found in insects, and that bears the genomic hallmarks of a new viral family. PMID:29340209

Bellerophon: A program to detect chimeric sequences in multiple sequence alignments

DOE Office of Scientific and Technical Information (OSTI.GOV)

Huber, Thomas; Faulkner, Geoffrey; Hugenholtz, Philip

2003-12-23

Bellerophon is a program for detecting chimeric sequences in multiple sequence datasets by an adaption of partial treeing analysis. Bellerophon was specifically developed to detect 16S rRNA gene chimeras in PCR-clone libraries of environmental samples but can be applied to other nucleotide sequence alignments.

Identification of Prostate Cancer-Specific microDNAs

DTIC Science & Technology

2016-02-01

circular DNA by rolling circle amplification (RCA) and then amplified DNA fragments were subject to deep sequencing. Deep sequencing of the...demonstrate the existence of microDNAs in prostate cancer. We adopted multiple displacement amplification (MDA) with random 2 primers for enriched...prostate cancer cells through multiple displacement amplification and next generation sequencing. R e la ti v e c e ll g ro w th ( % ) 0 20

Modeling the effects of diagenesis on carbonate clumped-isotope values in deep- and shallow-water settings

NASA Astrophysics Data System (ADS)

Stolper, Daniel A.; Eiler, John M.; Higgins, John A.

2018-04-01

The measurement of multiply isotopically substituted ('clumped isotope') carbonate groups provides a way to reconstruct past mineral formation temperatures. However, dissolution-reprecipitation (i.e., recrystallization) reactions, which commonly occur during sedimentary burial, can alter a sample's clumped-isotope composition such that it partially or wholly reflects deeper burial temperatures. Here we derive a quantitative model of diagenesis to explore how diagenesis alters carbonate clumped-isotope values. We apply the model to a new dataset from deep-sea sediments taken from Ocean Drilling Project site 807 in the equatorial Pacific. This dataset is used to ground truth the model. We demonstrate that the use of the model with accompanying carbonate clumped-isotope and carbonate δ18O values provides new constraints on both the diagenetic history of deep-sea settings as well as past equatorial sea-surface temperatures. Specifically, the combination of the diagenetic model and data support previous work that indicates equatorial sea-surface temperatures were warmer in the Paleogene as compared to today. We then explore whether the model is applicable to shallow-water settings commonly preserved in the rock record. Using a previously published dataset from the Bahamas, we demonstrate that the model captures the main trends of the data as a function of burial depth and thus appears applicable to a range of depositional settings.

Sequence-specific bias correction for RNA-seq data using recurrent neural networks.

PubMed

Zhang, Yao-Zhong; Yamaguchi, Rui; Imoto, Seiya; Miyano, Satoru

2017-01-25

The recent success of deep learning techniques in machine learning and artificial intelligence has stimulated a great deal of interest among bioinformaticians, who now wish to bring the power of deep learning to bare on a host of bioinformatical problems. Deep learning is ideally suited for biological problems that require automatic or hierarchical feature representation for biological data when prior knowledge is limited. In this work, we address the sequence-specific bias correction problem for RNA-seq data redusing Recurrent Neural Networks (RNNs) to model nucleotide sequences without pre-determining sequence structures. The sequence-specific bias of a read is then calculated based on the sequence probabilities estimated by RNNs, and used in the estimation of gene abundance. We explore the application of two popular RNN recurrent units for this task and demonstrate that RNN-based approaches provide a flexible way to model nucleotide sequences without knowledge of predetermined sequence structures. Our experiments show that training a RNN-based nucleotide sequence model is efficient and RNN-based bias correction methods compare well with the-state-of-the-art sequence-specific bias correction method on the commonly used MAQC-III data set. RNNs provides an alternative and flexible way to calculate sequence-specific bias without explicitly pre-determining sequence structures.

XPAT: a toolkit to conduct cross-platform association studies with heterogeneous sequencing datasets.

PubMed

Yu, Yao; Hu, Hao; Bohlender, Ryan J; Hu, Fulan; Chen, Jiun-Sheng; Holt, Carson; Fowler, Jerry; Guthery, Stephen L; Scheet, Paul; Hildebrandt, Michelle A T; Yandell, Mark; Huff, Chad D

2018-04-06

High-throughput sequencing data are increasingly being made available to the research community for secondary analyses, providing new opportunities for large-scale association studies. However, heterogeneity in target capture and sequencing technologies often introduce strong technological stratification biases that overwhelm subtle signals of association in studies of complex traits. Here, we introduce the Cross-Platform Association Toolkit, XPAT, which provides a suite of tools designed to support and conduct large-scale association studies with heterogeneous sequencing datasets. XPAT includes tools to support cross-platform aware variant calling, quality control filtering, gene-based association testing and rare variant effect size estimation. To evaluate the performance of XPAT, we conducted case-control association studies for three diseases, including 783 breast cancer cases, 272 ovarian cancer cases, 205 Crohn disease cases and 3507 shared controls (including 1722 females) using sequencing data from multiple sources. XPAT greatly reduced Type I error inflation in the case-control analyses, while replicating many previously identified disease-gene associations. We also show that association tests conducted with XPAT using cross-platform data have comparable performance to tests using matched platform data. XPAT enables new association studies that combine existing sequencing datasets to identify genetic loci associated with common diseases and other complex traits.

VaDiR: an integrated approach to Variant Detection in RNA.

PubMed

Neums, Lisa; Suenaga, Seiji; Beyerlein, Peter; Anders, Sara; Koestler, Devin; Mariani, Andrea; Chien, Jeremy

2018-02-01

Advances in next-generation DNA sequencing technologies are now enabling detailed characterization of sequence variations in cancer genomes. With whole-genome sequencing, variations in coding and non-coding sequences can be discovered. But the cost associated with it is currently limiting its general use in research. Whole-exome sequencing is used to characterize sequence variations in coding regions, but the cost associated with capture reagents and biases in capture rate limit its full use in research. Additional limitations include uncertainty in assigning the functional significance of the mutations when these mutations are observed in the non-coding region or in genes that are not expressed in cancer tissue. We investigated the feasibility of uncovering mutations from expressed genes using RNA sequencing datasets with a method called Variant Detection in RNA(VaDiR) that integrates 3 variant callers, namely: SNPiR, RVBoost, and MuTect2. The combination of all 3 methods, which we called Tier 1 variants, produced the highest precision with true positive mutations from RNA-seq that could be validated at the DNA level. We also found that the integration of Tier 1 variants with those called by MuTect2 and SNPiR produced the highest recall with acceptable precision. Finally, we observed a higher rate of mutation discovery in genes that are expressed at higher levels. Our method, VaDiR, provides a possibility of uncovering mutations from RNA sequencing datasets that could be useful in further functional analysis. In addition, our approach allows orthogonal validation of DNA-based mutation discovery by providing complementary sequence variation analysis from paired RNA/DNA sequencing datasets.

Comparison Between NPP-VIIRS Aerosol Data Products and the MODIS AQUA Deep Blue Collection 6 Dataset Over Land

NASA Technical Reports Server (NTRS)

Sayer, Andrew M.; Hsu, N. C.; Bettenhausen, C.; Lee, J.; Kondragunta, S.

2013-01-01

Aerosols are small particles suspended in the atmosphere and have a variety of natural and man-made sources. Knowledge of aerosol optical depth (AOD), which is a measure of the amount of aerosol in the atmosphere, and its change over time, is important for multiple reasons. These include climate change, air quality (pollution) monitoring, monitoring hazards such as dust storms and volcanic ash, monitoring smoke from biomass burning, determining potential energy yields from solar plants, determining visibility at sea, estimating fertilization of oceans and rainforests by transported mineral dust, understanding changes in weather brought upon by the interaction of aerosols and clouds, and more. The Suomi-NPP satellite was launched late in 2011. The Visible Infrared Imaging Radiometer Suite (VIIRS) aboard Suomi-NPP is being used, among other things, to determine AOD. This study compares the VIIRS dataset to ground-based measurements of AOD, along with a state-of-the-art satellite AOD dataset (the new version of the Moderate Resolution Imaging Spectrometer Deep Blue algorithm) to assess its reliability. The Suomi-NPP satellite was launched late in 2011, carrying several instruments designed to continue the biogeophysical data records of current and previous satellite sensors. The Visible Infrared Imaging Radiometer Suite (VIIRS) aboard Suomi-NPP is being used, among other things, to determine aerosol optical depth (AOD), and related activities since launch have been focused towards validating and understanding this new dataset through comparisons with other satellite and ground-based products. The operational VIIRS AOD product is compared over land with AOD derived from Moderate Resolution Imaging Spectrometer (MODIS) observations using the Deep Blue (DB) algorithm from the forthcoming Collection 6 of MODIS data

Deep learning in breast cancer risk assessment: evaluation of fine-tuned convolutional neural networks on a clinical dataset of FFDMs

NASA Astrophysics Data System (ADS)

Li, Hui; Mendel, Kayla R.; Lee, John H.; Lan, Li; Giger, Maryellen L.

2018-02-01

We evaluated the potential of deep learning in the assessment of breast cancer risk using convolutional neural networks (CNNs) fine-tuned on full-field digital mammographic (FFDM) images. This study included 456 clinical FFDM cases from two high-risk datasets: BRCA1/2 gene-mutation carriers (53 cases) and unilateral cancer patients (75 cases), and a low-risk dataset as the control group (328 cases). All FFDM images (12-bit quantization and 100 micron pixel) were acquired with a GE Senographe 2000D system and were retrospectively collected under an IRB-approved, HIPAA-compliant protocol. Regions of interest of 256x256 pixels were selected from the central breast region behind the nipple in the craniocaudal projection. VGG19 pre-trained on the ImageNet dataset was used to classify the images either as high-risk or as low-risk subjects. The last fully-connected layer of pre-trained VGG19 was fine-tuned on FFDM images for breast cancer risk assessment. Performance was evaluated using the area under the receiver operating characteristic (ROC) curve (AUC) in the task of distinguishing between high-risk and low-risk subjects. AUC values of 0.84 (SE=0.05) and 0.72 (SE=0.06) were obtained in the task of distinguishing between the BRCA1/2 gene-mutation carriers and low-risk women and between unilateral cancer patients and low-risk women, respectively. Deep learning with CNNs appears to be able to extract parenchymal characteristics directly from FFDMs which are relevant to the task of distinguishing between cancer risk populations, and therefore has potential to aid clinicians in assessing mammographic parenchymal patterns for cancer risk assessment.

Protein remote homology detection based on bidirectional long short-term memory.

PubMed

Li, Shumin; Chen, Junjie; Liu, Bin

2017-10-10

Protein remote homology detection plays a vital role in studies of protein structures and functions. Almost all of the traditional machine leaning methods require fixed length features to represent the protein sequences. However, it is never an easy task to extract the discriminative features with limited knowledge of proteins. On the other hand, deep learning technique has demonstrated its advantage in automatically learning representations. It is worthwhile to explore the applications of deep learning techniques to the protein remote homology detection. In this study, we employ the Bidirectional Long Short-Term Memory (BLSTM) to learn effective features from pseudo proteins, also propose a predictor called ProDec-BLSTM: it includes input layer, bidirectional LSTM, time distributed dense layer and output layer. This neural network can automatically extract the discriminative features by using bidirectional LSTM and the time distributed dense layer. Experimental results on a widely-used benchmark dataset show that ProDec-BLSTM outperforms other related methods in terms of both the mean ROC and mean ROC50 scores. This promising result shows that ProDec-BLSTM is a useful tool for protein remote homology detection. Furthermore, the hidden patterns learnt by ProDec-BLSTM can be interpreted and visualized, and therefore, additional useful information can be obtained.

Human action classification using procrustes shape theory

NASA Astrophysics Data System (ADS)

Cho, Wanhyun; Kim, Sangkyoon; Park, Soonyoung; Lee, Myungeun

2015-02-01

In this paper, we propose new method that can classify a human action using Procrustes shape theory. First, we extract a pre-shape configuration vector of landmarks from each frame of an image sequence representing an arbitrary human action, and then we have derived the Procrustes fit vector for pre-shape configuration vector. Second, we extract a set of pre-shape vectors from tanning sample stored at database, and we compute a Procrustes mean shape vector for these preshape vectors. Third, we extract a sequence of the pre-shape vectors from input video, and we project this sequence of pre-shape vectors on the tangent space with respect to the pole taking as a sequence of mean shape vectors corresponding with a target video. And we calculate the Procrustes distance between two sequences of the projection pre-shape vectors on the tangent space and the mean shape vectors. Finally, we classify the input video into the human action class with minimum Procrustes distance. We assess a performance of the proposed method using one public dataset, namely Weizmann human action dataset. Experimental results reveal that the proposed method performs very good on this dataset.

Genomic Datasets for Cancer Research

Cancer.gov

A variety of datasets from genome-wide association studies of cancer and other genotype-phenotype studies, including sequencing and molecular diagnostic assays, are available to approved investigators through the Extramural National Cancer Institute Data Access Committee.

Multi-categorical deep learning neural network to classify retinal images: A pilot study employing small database

PubMed Central

Seo, Jeong Gi; Kwak, Jiyong; Um, Terry Taewoong; Rim, Tyler Hyungtaek

2017-01-01

Deep learning emerges as a powerful tool for analyzing medical images. Retinal disease detection by using computer-aided diagnosis from fundus image has emerged as a new method. We applied deep learning convolutional neural network by using MatConvNet for an automated detection of multiple retinal diseases with fundus photographs involved in STructured Analysis of the REtina (STARE) database. Dataset was built by expanding data on 10 categories, including normal retina and nine retinal diseases. The optimal outcomes were acquired by using a random forest transfer learning based on VGG-19 architecture. The classification results depended greatly on the number of categories. As the number of categories increased, the performance of deep learning models was diminished. When all 10 categories were included, we obtained results with an accuracy of 30.5%, relative classifier information (RCI) of 0.052, and Cohen’s kappa of 0.224. Considering three integrated normal, background diabetic retinopathy, and dry age-related macular degeneration, the multi-categorical classifier showed accuracy of 72.8%, 0.283 RCI, and 0.577 kappa. In addition, several ensemble classifiers enhanced the multi-categorical classification performance. The transfer learning incorporated with ensemble classifier of clustering and voting approach presented the best performance with accuracy of 36.7%, 0.053 RCI, and 0.225 kappa in the 10 retinal diseases classification problem. First, due to the small size of datasets, the deep learning techniques in this study were ineffective to be applied in clinics where numerous patients suffering from various types of retinal disorders visit for diagnosis and treatment. Second, we found that the transfer learning incorporated with ensemble classifiers can improve the classification performance in order to detect multi-categorical retinal diseases. Further studies should confirm the effectiveness of algorithms with large datasets obtained from hospitals. PMID:29095872

An Efficient Strategy of Screening for Pathogens in Wild-Caught Ticks and Mosquitoes by Reusing Small RNA Deep Sequencing Data

PubMed Central

An, Xiaoping; Fan, Hang; Ma, Maijuan; Anderson, Benjamin D.; Jiang, Jiafu; Liu, Wei; Cao, Wuchun; Tong, Yigang

2014-01-01

This paper explored our hypothesis that sRNA (18∼30 bp) deep sequencing technique can be used as an efficient strategy to identify microorganisms other than viruses, such as prokaryotic and eukaryotic pathogens. In the study, the clean reads derived from the sRNA deep sequencing data of wild-caught ticks and mosquitoes were compared against the NCBI nucleotide collection (non-redundant nt database) using Blastn. The blast results were then analyzed with in-house Python scripts. An empirical formula was proposed to identify the putative pathogens. Results showed that not only viruses but also prokaryotic and eukaryotic species of interest can be screened out and were subsequently confirmed with experiments. Specially, a novel Rickettsia spp. was indicated to exist in Haemaphysalis longicornis ticks collected in Beijing. Our study demonstrated the reuse of sRNA deep sequencing data would have the potential to trace the origin of pathogens or discover novel agents of emerging/re-emerging infectious diseases. PMID:24618575

Draft Genome Sequence of Deep-Sea Alteromonas sp. Strain V450 Isolated from the Marine Sponge Leiodermatium sp.

PubMed

Wang, Guojun; Barrett, Nolan H; McCarthy, Peter J

2017-02-02

The proteobacterium Alteromonas sp. strain V450 was isolated from the Atlantic deep-sea sponge Leiodermatium sp. Here, we report the draft genome sequence of this strain, with a genome size of approx. 4.39 Mb and a G+C content of 44.01%. The results will aid deep-sea microbial ecology, evolution, and sponge-microbe association studies. Copyright © 2017 Wang et al.

Evaluation of a Traffic Sign Detector by Synthetic Image Data for Advanced Driver Assistance Systems

NASA Astrophysics Data System (ADS)

Hanel, A.; Kreuzpaintner, D.; Stilla, U.

2018-05-01

Recently, several synthetic image datasets of street scenes have been published. These datasets contain various traffic signs and can therefore be used to train and test machine learning-based traffic sign detectors. In this contribution, selected datasets are compared regarding ther applicability for traffic sign detection. The comparison covers the process to produce the synthetic images and addresses the virtual worlds, needed to produce the synthetic images, and their environmental conditions. The comparison covers variations in the appearance of traffic signs and the labeling strategies used for the datasets, as well. A deep learning traffic sign detector is trained with multiple training datasets with different ratios between synthetic and real training samples to evaluate the synthetic SYNTHIA dataset. A test of the detector on real samples only has shown that an overall accuracy and ROC AUC of more than 95 % can be achieved for both a small rate of synthetic samples and a large rate of synthetic samples in the training dataset.

Volumetric multimodality neural network for brain tumor segmentation

NASA Astrophysics Data System (ADS)

Silvana Castillo, Laura; Alexandra Daza, Laura; Carlos Rivera, Luis; Arbeláez, Pablo

2017-11-01

Brain lesion segmentation is one of the hardest tasks to be solved in computer vision with an emphasis on the medical field. We present a convolutional neural network that produces a semantic segmentation of brain tumors, capable of processing volumetric data along with information from multiple MRI modalities at the same time. This results in the ability to learn from small training datasets and highly imbalanced data. Our method is based on DeepMedic, the state of the art in brain lesion segmentation. We develop a new architecture with more convolutional layers, organized in three parallel pathways with different input resolution, and additional fully connected layers. We tested our method over the 2015 BraTS Challenge dataset, reaching an average dice coefficient of 84%, while the standard DeepMedic implementation reached 74%.

An adaptive deep learning approach for PPG-based identification.

PubMed

Jindal, V; Birjandtalab, J; Pouyan, M Baran; Nourani, M

2016-08-01

Wearable biosensors have become increasingly popular in healthcare due to their capabilities for low cost and long term biosignal monitoring. This paper presents a novel two-stage technique to offer biometric identification using these biosensors through Deep Belief Networks and Restricted Boltzman Machines. Our identification approach improves robustness in current monitoring procedures within clinical, e-health and fitness environments using Photoplethysmography (PPG) signals through deep learning classification models. The approach is tested on TROIKA dataset using 10-fold cross validation and achieved an accuracy of 96.1%.

miRBase: integrating microRNA annotation and deep-sequencing data.

PubMed

Kozomara, Ana; Griffiths-Jones, Sam

2011-01-01

miRBase is the primary online repository for all microRNA sequences and annotation. The current release (miRBase 16) contains over 15,000 microRNA gene loci in over 140 species, and over 17,000 distinct mature microRNA sequences. Deep-sequencing technologies have delivered a sharp rise in the rate of novel microRNA discovery. We have mapped reads from short RNA deep-sequencing experiments to microRNAs in miRBase and developed web interfaces to view these mappings. The user can view all read data associated with a given microRNA annotation, filter reads by experiment and count, and search for microRNAs by tissue- and stage-specific expression. These data can be used as a proxy for relative expression levels of microRNA sequences, provide detailed evidence for microRNA annotations and alternative isoforms of mature microRNAs, and allow us to revisit previous annotations. miRBase is available online at: http://www.mirbase.org/.

Transcriptome sequences resolve deep relationships of the grape family.

PubMed

Wen, Jun; Xiong, Zhiqiang; Nie, Ze-Long; Mao, Likai; Zhu, Yabing; Kan, Xian-Zhao; Ickert-Bond, Stefanie M; Gerrath, Jean; Zimmer, Elizabeth A; Fang, Xiao-Dong

2013-01-01

Previous phylogenetic studies of the grape family (Vitaceae) yielded poorly resolved deep relationships, thus impeding our understanding of the evolution of the family. Next-generation sequencing now offers access to protein coding sequences very easily, quickly and cost-effectively. To improve upon earlier work, we extracted 417 orthologous single-copy nuclear genes from the transcriptomes of 15 species of the Vitaceae, covering its phylogenetic diversity. The resulting transcriptome phylogeny provides robust support for the deep relationships, showing the phylogenetic utility of transcriptome data for plants over a time scale at least since the mid-Cretaceous. The pros and cons of transcriptome data for phylogenetic inference in plants are also evaluated.

Adaptive Neuron Apoptosis for Accelerating Deep Learning on Large Scale Systems

DOE Office of Scientific and Technical Information (OSTI.GOV)

Siegel, Charles M.; Daily, Jeffrey A.; Vishnu, Abhinav

Machine Learning and Data Mining (MLDM) algorithms are becoming ubiquitous in {\\em model learning} from the large volume of data generated using simulations, experiments and handheld devices. Deep Learning algorithms -- a class of MLDM algorithms -- are applied for automatic feature extraction, and learning non-linear models for unsupervised and supervised algorithms. Naturally, several libraries which support large scale Deep Learning -- such as TensorFlow and Caffe -- have become popular. In this paper, we present novel techniques to accelerate the convergence of Deep Learning algorithms by conducting low overhead removal of redundant neurons -- {\\em apoptosis} of neurons --more » which do not contribute to model learning, during the training phase itself. We provide in-depth theoretical underpinnings of our heuristics (bounding accuracy loss and handling apoptosis of several neuron types), and present the methods to conduct adaptive neuron apoptosis. We implement our proposed heuristics with the recently introduced TensorFlow and using its recently proposed extension with MPI. Our performance evaluation on two difference clusters -- one connected with Intel Haswell multi-core systems, and other with nVIDIA GPUs -- using InfiniBand, indicates the efficacy of the proposed heuristics and implementations. Specifically, we are able to improve the training time for several datasets by 2-3x, while reducing the number of parameters by 30x (4-5x on average) on datasets such as ImageNet classification. For the Higgs Boson dataset, our implementation improves the accuracy (measured by Area Under Curve (AUC)) for classification from 0.88/1 to 0.94/1, while reducing the number of parameters by 3x in comparison to existing literature, while achieving a 2.44x speedup in comparison to the default (no apoptosis) algorithm.« less

The Livermore Brain: Massive Deep Learning Networks Enabled by High Performance Computing

DOE Office of Scientific and Technical Information (OSTI.GOV)

Chen, Barry Y.

The proliferation of inexpensive sensor technologies like the ubiquitous digital image sensors has resulted in the collection and sharing of vast amounts of unsorted and unexploited raw data. Companies and governments who are able to collect and make sense of large datasets to help them make better decisions more rapidly will have a competitive advantage in the information era. Machine Learning technologies play a critical role for automating the data understanding process; however, to be maximally effective, useful intermediate representations of the data are required. These representations or “features” are transformations of the raw data into a form where patternsmore » are more easily recognized. Recent breakthroughs in Deep Learning have made it possible to learn these features from large amounts of labeled data. The focus of this project is to develop and extend Deep Learning algorithms for learning features from vast amounts of unlabeled data and to develop the HPC neural network training platform to support the training of massive network models. This LDRD project succeeded in developing new unsupervised feature learning algorithms for images and video and created a scalable neural network training toolkit for HPC. Additionally, this LDRD helped create the world’s largest freely-available image and video dataset supporting open multimedia research and used this dataset for training our deep neural networks. This research helped LLNL capture several work-for-others (WFO) projects, attract new talent, and establish collaborations with leading academic and commercial partners. Finally, this project demonstrated the successful training of the largest unsupervised image neural network using HPC resources and helped establish LLNL leadership at the intersection of Machine Learning and HPC research.« less

Deep ensemble learning of virtual endoluminal views for polyp detection in CT colonography

NASA Astrophysics Data System (ADS)

Umehara, Kensuke; Näppi, Janne J.; Hironaka, Toru; Regge, Daniele; Ishida, Takayuki; Yoshida, Hiroyuki

2017-03-01

Robust training of a deep convolutional neural network (DCNN) requires a very large number of annotated datasets that are currently not available in CT colonography (CTC). We previously demonstrated that deep transfer learning provides an effective approach for robust application of a DCNN in CTC. However, at high detection accuracy, the differentiation of small polyps from non-polyps was still challenging. In this study, we developed and evaluated a deep ensemble learning (DEL) scheme for reviewing of virtual endoluminal images to improve the performance of computer-aided detection (CADe) of polyps in CTC. Nine different types of image renderings were generated from virtual endoluminal images of polyp candidates detected by a conventional CADe system. Eleven DCNNs that represented three types of publically available pre-trained DCNN models were re-trained by transfer learning to identify polyps from the virtual endoluminal images. A DEL scheme that determines the final detected polyps by a review of the nine types of VE images was developed by combining the DCNNs using a random forest classifier as a meta-classifier. For evaluation, we sampled 154 CTC cases from a large CTC screening trial and divided the cases randomly into a training dataset and a test dataset. At 3.9 falsepositive (FP) detections per patient on average, the detection sensitivities of the conventional CADe system, the highestperforming single DCNN, and the DEL scheme were 81.3%, 90.7%, and 93.5%, respectively, for polyps ≥6 mm in size. For small polyps, the DEL scheme reduced the number of false positives by up to 83% over that of using a single DCNN alone. These preliminary results indicate that the DEL scheme provides an effective approach for improving the polyp detection performance of CADe in CTC, especially for small polyps.

Assessing the utility of transcriptome data for inferring phylogenetic relationships among coleoid cephalopods.

PubMed

Lindgren, Annie R; Anderson, Frank E

2018-01-01

Historically, deep-level relationships within the molluscan class Cephalopoda (squids, cuttlefishes, octopods and their relatives) have remained elusive due in part to the considerable morphological diversity of extant taxa, a limited fossil record for species that lack a calcareous shell and difficulties in sampling open ocean taxa. Many conflicts identified by morphologists in the early 1900s remain unresolved today in spite of advances in morphological, molecular and analytical methods. In this study we assess the utility of transcriptome data for resolving cephalopod phylogeny, with special focus on the orders of Decapodiformes (open-eye squids, bobtail squids, cuttlefishes and relatives). To do so, we took new and previously published transcriptome data and used a unique cephalopod core ortholog set to generate a dataset that was subjected to an array of filtering and analytical methods to assess the impacts of: taxon sampling, ortholog number, compositional and rate heterogeneity and incongruence across loci. Analyses indicated that datasets that maximized taxonomic coverage but included fewer orthologs were less stable than datasets that sacrificed taxon sampling to increase the number of orthologs. Clades recovered irrespective of dataset, filtering or analytical method included Octopodiformes (Vampyroteuthis infernalis + octopods), Decapodiformes (squids, cuttlefishes and their relatives), and orders Oegopsida (open-eyed squids) and Myopsida (e.g., loliginid squids). Ordinal-level relationships within Decapodiformes were the most susceptible to dataset perturbation, further emphasizing the challenges associated with uncovering relationships at deep nodes in the cephalopod tree of life. Copyright © 2017 Elsevier Inc. All rights reserved.

Diverse and Abundant Secondary Metabolism Biosynthetic Gene Clusters in the Genomes of Marine Sponge Derived Streptomyces spp. Isolates.

PubMed

Jackson, Stephen A; Crossman, Lisa; Almeida, Eduardo L; Margassery, Lekha Menon; Kennedy, Jonathan; Dobson, Alan D W

2018-02-20

The genus Streptomyces produces secondary metabolic compounds that are rich in biological activity. Many of these compounds are genetically encoded by large secondary metabolism biosynthetic gene clusters (smBGCs) such as polyketide synthases (PKS) and non-ribosomal peptide synthetases (NRPS) which are modular and can be highly repetitive. Due to the repeats, these gene clusters can be difficult to resolve using short read next generation datasets and are often quite poorly predicted using standard approaches. We have sequenced the genomes of 13 Streptomyces spp. strains isolated from shallow water and deep-sea sponges that display antimicrobial activities against a number of clinically relevant bacterial and yeast species. Draft genomes have been assembled and smBGCs have been identified using the antiSMASH (antibiotics and Secondary Metabolite Analysis Shell) web platform. We have compared the smBGCs amongst strains in the search for novel sequences conferring the potential to produce novel bioactive secondary metabolites. The strains in this study recruit to four distinct clades within the genus Streptomyces . The marine strains host abundant smBGCs which encode polyketides, NRPS, siderophores, bacteriocins and lantipeptides. The deep-sea strains appear to be enriched with gene clusters encoding NRPS. Marine adaptations are evident in the sponge-derived strains which are enriched for genes involved in the biosynthesis and transport of compatible solutes and for heat-shock proteins. Streptomyces spp. from marine environments are a promising source of novel bioactive secondary metabolites as the abundance and diversity of smBGCs show high degrees of novelty. Sponge derived Streptomyces spp. isolates appear to display genomic adaptations to marine living when compared to terrestrial strains.

Identification of characteristic oligonucleotides in the bacterial 16S ribosomal RNA sequence dataset

NASA Technical Reports Server (NTRS)

Zhang, Zhengdong; Willson, Richard C.; Fox, George E.

2002-01-01

MOTIVATION: The phylogenetic structure of the bacterial world has been intensively studied by comparing sequences of 16S ribosomal RNA (16S rRNA). This database of sequences is now widely used to design probes for the detection of specific bacteria or groups of bacteria one at a time. The success of such methods reflects the fact that there are local sequence segments that are highly characteristic of particular organisms or groups of organisms. It is not clear, however, the extent to which such signature sequences exist in the 16S rRNA dataset. A better understanding of the numbers and distribution of highly informative oligonucleotide sequences may facilitate the design of hybridization arrays that can characterize the phylogenetic position of an unknown organism or serve as the basis for the development of novel approaches for use in bacterial identification. RESULTS: A computer-based algorithm that characterizes the extent to which any individual oligonucleotide sequence in 16S rRNA is characteristic of any particular bacterial grouping was developed. A measure of signature quality, Q(s), was formulated and subsequently calculated for every individual oligonucleotide sequence in the size range of 5-11 nucleotides and for 15mers with reference to each cluster and subcluster in a 929 organism representative phylogenetic tree. Subsequently, the perfect signature sequences were compared to the full set of 7322 sequences to see how common false positives were. The work completed here establishes beyond any doubt that highly characteristic oligonucleotides exist in the bacterial 16S rRNA sequence dataset in large numbers. Over 16,000 15mers were identified that might be useful as signatures. Signature oligonucleotides are available for over 80% of the nodes in the representative tree.

Simultaneous bright‐ and black‐blood whole‐heart MRI for noncontrast enhanced coronary lumen and thrombus visualization

PubMed Central

Neji, Radhouene; Phinikaridou, Alkystis; Whitaker, John; Botnar, René M.; Prieto, Claudia

2017-01-01

Purpose To develop a 3D whole‐heart Bright‐blood and black‐blOOd phase SensiTive (BOOST) inversion recovery sequence for simultaneous noncontrast enhanced coronary lumen and thrombus/hemorrhage visualization. Methods The proposed sequence alternates the acquisition of two bright‐blood datasets preceded by different preparatory pulses to obtain variations in blood/myocardium contrast, which then are combined in a phase‐sensitive inversion recovery (PSIR)‐like reconstruction to obtain a third, coregistered, black‐blood dataset. The bright‐blood datasets are used for both visualization of the coronary lumen and motion estimation, whereas the complementary black‐blood dataset potentially allows for thrombus/hemorrhage visualization. Furthermore, integration with 2D image‐based navigation enables 100% scan efficiency and predictable scan times. The proposed sequence was compared to conventional coronary MR angiography (CMRA) and PSIR sequences in a standardized phantom and in healthy subjects. Feasibility for thrombus depiction was tested ex vivo. Results With BOOST, the coronary lumen is visualized with significantly higher (P < 0.05) contrast‐to‐noise ratio and vessel sharpness when compared to conventional CMRA. Furthermore, BOOST showed effective blood signal suppression as well as feasibility for thrombus visualization ex vivo. Conclusion A new PSIR sequence for noncontrast enhanced simultaneous coronary lumen and thrombus/hemorrhage detection was developed. The sequence provided improved coronary lumen depiction and showed potential for thrombus visualization. Magn Reson Med 79:1460–1472, 2018. © 2017 International Society for Magnetic Resonance in Medicine. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. PMID:28722267

Draft Genome Sequence of Pseudomonas oceani DSM 100277T, a Deep-Sea Bacterium

PubMed Central

2018-01-01

ABSTRACT Pseudomonas oceani DSM 100277T was isolated from deep seawater in the Okinawa Trough at 1390 m. P. oceani belongs to the Pseudomonas pertucinogena group. Here, we report the draft genome sequence of P. oceani, which has an estimated size of 4.1 Mb and exhibits 3,790 coding sequences, with a G+C content of 59.94 mol%. PMID:29650573

A Bioinformatic Pipeline for Monitoring of the Mutational Stability of Viral Drug Targets with Deep-Sequencing Technology.

PubMed

Kravatsky, Yuri; Chechetkin, Vladimir; Fedoseeva, Daria; Gorbacheva, Maria; Kravatskaya, Galina; Kretova, Olga; Tchurikov, Nickolai

2017-11-23

The efficient development of antiviral drugs, including efficient antiviral small interfering RNAs (siRNAs), requires continuous monitoring of the strict correspondence between a drug and the related highly variable viral DNA/RNA target(s). Deep sequencing is able to provide an assessment of both the general target conservation and the frequency of particular mutations in the different target sites. The aim of this study was to develop a reliable bioinformatic pipeline for the analysis of millions of short, deep sequencing reads corresponding to selected highly variable viral sequences that are drug target(s). The suggested bioinformatic pipeline combines the available programs and the ad hoc scripts based on an original algorithm of the search for the conserved targets in the deep sequencing data. We also present the statistical criteria for the threshold of reliable mutation detection and for the assessment of variations between corresponding data sets. These criteria are robust against the possible sequencing errors in the reads. As an example, the bioinformatic pipeline is applied to the study of the conservation of RNA interference (RNAi) targets in human immunodeficiency virus 1 (HIV-1) subtype A. The developed pipeline is freely available to download at the website http://virmut.eimb.ru/. Brief comments and comparisons between VirMut and other pipelines are also presented.

Deep Motif Dashboard: Visualizing and Understanding Genomic Sequences Using Deep Neural Networks

PubMed Central

Lanchantin, Jack; Singh, Ritambhara; Wang, Beilun; Qi, Yanjun

2018-01-01

Deep neural network (DNN) models have recently obtained state-of-the-art prediction accuracy for the transcription factor binding (TFBS) site classification task. However, it remains unclear how these approaches identify meaningful DNA sequence signals and give insights as to why TFs bind to certain locations. In this paper, we propose a toolkit called the Deep Motif Dashboard (DeMo Dashboard) which provides a suite of visualization strategies to extract motifs, or sequence patterns from deep neural network models for TFBS classification. We demonstrate how to visualize and understand three important DNN models: convolutional, recurrent, and convolutional-recurrent networks. Our first visualization method is finding a test sequence’s saliency map which uses first-order derivatives to describe the importance of each nucleotide in making the final prediction. Second, considering recurrent models make predictions in a temporal manner (from one end of a TFBS sequence to the other), we introduce temporal output scores, indicating the prediction score of a model over time for a sequential input. Lastly, a class-specific visualization strategy finds the optimal input sequence for a given TFBS positive class via stochastic gradient optimization. Our experimental results indicate that a convolutional-recurrent architecture performs the best among the three architectures. The visualization techniques indicate that CNN-RNN makes predictions by modeling both motifs as well as dependencies among them. PMID:27896980

Aftershock occurrence rate decay for individual sequences and catalogs

NASA Astrophysics Data System (ADS)

Nyffenegger, Paul A.

One of the earliest observations of the Earth's seismicity is that the rate of aftershock occurrence decays with time according to a power law commonly known as modified Omori-law (MOL) decay. However, the physical reasons for aftershock occurrence and the empirical decay in rate remain unclear despite numerous models that yield similar rate decay behavior. Key problems in relating the observed empirical relationship to the physical conditions of the mainshock and fault are the lack of studies including small magnitude mainshocks and the lack of uniformity between studies. We use simulated aftershock sequences to investigate the factors which influence the maximum likelihood (ML) estimate of the Omori-law p value, the parameter describing aftershock occurrence rate decay, for both individual aftershock sequences and "stacked" or superposed sequences. Generally the ML estimate of p is accurate, but since the ML estimated uncertainty is unaffected by whether the sequence resembles an MOL model, a goodness-of-fit test such as the Anderson-Darling statistic is necessary. While stacking aftershock sequences permits the study of entire catalogs and sequences with small aftershock populations, stacking introduces artifacts. The p value for stacked sequences is approximately equal to the mean of the individual sequence p values. We apply single-link cluster analysis to identify all aftershock sequences from eleven regional seismicity catalogs. We observe two new mathematically predictable empirical relationships for the distribution of aftershock sequence populations. The average properties of aftershock sequences are not correlated with tectonic environment, but aftershock populations and p values do show a depth dependence. The p values show great variability with time, and large values or changes in p sometimes precedes major earthquakes. Studies of teleseismic earthquake catalogs over the last twenty years have led seismologists to question seismicity models and aftershock sequence decay for deep sequences. For seven exceptional deep sequences, we conclude that MOL decay adequately describes these sequences, and little difference exists compared to shallow sequences. However, they do include larger aftershock populations compared to most deep sequences. These results imply that p values for deep sequences are larger than those for intermediate depth sequences.

A high level interface to SCOP and ASTRAL implemented in python.

PubMed

Casbon, James A; Crooks, Gavin E; Saqi, Mansoor A S

2006-01-10

Benchmarking algorithms in structural bioinformatics often involves the construction of datasets of proteins with given sequence and structural properties. The SCOP database is a manually curated structural classification which groups together proteins on the basis of structural similarity. The ASTRAL compendium provides non redundant subsets of SCOP domains on the basis of sequence similarity such that no two domains in a given subset share more than a defined degree of sequence similarity. Taken together these two resources provide a 'ground truth' for assessing structural bioinformatics algorithms. We present a small and easy to use API written in python to enable construction of datasets from these resources. We have designed a set of python modules to provide an abstraction of the SCOP and ASTRAL databases. The modules are designed to work as part of the Biopython distribution. Python users can now manipulate and use the SCOP hierarchy from within python programs, and use ASTRAL to return sequences of domains in SCOP, as well as clustered representations of SCOP from ASTRAL. The modules make the analysis and generation of datasets for use in structural genomics easier and more principled.

Study of infectious diseases in archaeological bone material - A dataset.

PubMed

Pucu, Elisa; Cascardo, Paula; Chame, Marcia; Felice, Gisele; Guidon, Niéde; Cleonice Vergne, Maria; Campos, Guadalupe; Roberto Machado-Silva, José; Leles, Daniela

2017-08-01

Bones of human and ground sloth remains were analyzed for presence of Trypanosoma cruzi by conventional PCR using primers TC, TC1 and TC2. Sequence results amplified a fragment with the same product size as the primers (300 and 350pb). Amplified PCR product was sequenced and analyzed on GenBank, using Blast. Although these sequences did not match with these parasites they showed high amplification with species of bacteria. This article presents the methodology used and the alignment of the sequences. The display of this dataset will allow further analysis of our results and discussion presented in the manuscript "Finding the unexpected: a critical view on molecular diagnosis of infectious diseases in archaeological samples" (Pucu et al. 2017) [1].

Protein Secondary Structure Prediction Using Deep Convolutional Neural Fields.

PubMed

Wang, Sheng; Peng, Jian; Ma, Jianzhu; Xu, Jinbo

2016-01-11

Protein secondary structure (SS) prediction is important for studying protein structure and function. When only the sequence (profile) information is used as input feature, currently the best predictors can obtain ~80% Q3 accuracy, which has not been improved in the past decade. Here we present DeepCNF (Deep Convolutional Neural Fields) for protein SS prediction. DeepCNF is a Deep Learning extension of Conditional Neural Fields (CNF), which is an integration of Conditional Random Fields (CRF) and shallow neural networks. DeepCNF can model not only complex sequence-structure relationship by a deep hierarchical architecture, but also interdependency between adjacent SS labels, so it is much more powerful than CNF. Experimental results show that DeepCNF can obtain ~84% Q3 accuracy, ~85% SOV score, and ~72% Q8 accuracy, respectively, on the CASP and CAMEO test proteins, greatly outperforming currently popular predictors. As a general framework, DeepCNF can be used to predict other protein structure properties such as contact number, disorder regions, and solvent accessibility.

Protein Secondary Structure Prediction Using Deep Convolutional Neural Fields

NASA Astrophysics Data System (ADS)

Wang, Sheng; Peng, Jian; Ma, Jianzhu; Xu, Jinbo

2016-01-01

Protein secondary structure (SS) prediction is important for studying protein structure and function. When only the sequence (profile) information is used as input feature, currently the best predictors can obtain ~80% Q3 accuracy, which has not been improved in the past decade. Here we present DeepCNF (Deep Convolutional Neural Fields) for protein SS prediction. DeepCNF is a Deep Learning extension of Conditional Neural Fields (CNF), which is an integration of Conditional Random Fields (CRF) and shallow neural networks. DeepCNF can model not only complex sequence-structure relationship by a deep hierarchical architecture, but also interdependency between adjacent SS labels, so it is much more powerful than CNF. Experimental results show that DeepCNF can obtain ~84% Q3 accuracy, ~85% SOV score, and ~72% Q8 accuracy, respectively, on the CASP and CAMEO test proteins, greatly outperforming currently popular predictors. As a general framework, DeepCNF can be used to predict other protein structure properties such as contact number, disorder regions, and solvent accessibility.

High precision in protein contact prediction using fully convolutional neural networks and minimal sequence features.

PubMed

Jones, David T; Kandathil, Shaun M

2018-04-26

In addition to substitution frequency data from protein sequence alignments, many state-of-the-art methods for contact prediction rely on additional sources of information, or features, of protein sequences in order to predict residue-residue contacts, such as solvent accessibility, predicted secondary structure, and scores from other contact prediction methods. It is unclear how much of this information is needed to achieve state-of-the-art results. Here, we show that using deep neural network models, simple alignment statistics contain sufficient information to achieve state-of-the-art precision. Our prediction method, DeepCov, uses fully convolutional neural networks operating on amino-acid pair frequency or covariance data derived directly from sequence alignments, without using global statistical methods such as sparse inverse covariance or pseudolikelihood estimation. Comparisons against CCMpred and MetaPSICOV2 show that using pairwise covariance data calculated from raw alignments as input allows us to match or exceed the performance of both of these methods. Almost all of the achieved precision is obtained when considering relatively local windows (around 15 residues) around any member of a given residue pairing; larger window sizes have comparable performance. Assessment on a set of shallow sequence alignments (fewer than 160 effective sequences) indicates that the new method is substantially more precise than CCMpred and MetaPSICOV2 in this regime, suggesting that improved precision is attainable on smaller sequence families. Overall, the performance of DeepCov is competitive with the state of the art, and our results demonstrate that global models, which employ features from all parts of the input alignment when predicting individual contacts, are not strictly needed in order to attain precise contact predictions. DeepCov is freely available at https://github.com/psipred/DeepCov. d.t.jones@ucl.ac.uk.

LS³: A Method for Improving Phylogenomic Inferences When Evolutionary Rates Are Heterogeneous among Taxa

PubMed Central

Rivera-Rivera, Carlos J.; Montoya-Burgos, Juan I.

2016-01-01

Phylogenetic inference artifacts can occur when sequence evolution deviates from assumptions made by the models used to analyze them. The combination of strong model assumption violations and highly heterogeneous lineage evolutionary rates can become problematic in phylogenetic inference, and lead to the well-described long-branch attraction (LBA) artifact. Here, we define an objective criterion for assessing lineage evolutionary rate heterogeneity among predefined lineages: the result of a likelihood ratio test between a model in which the lineages evolve at the same rate (homogeneous model) and a model in which different lineage rates are allowed (heterogeneous model). We implement this criterion in the algorithm Locus Specific Sequence Subsampling (LS³), aimed at reducing the effects of LBA in multi-gene datasets. For each gene, LS³ sequentially removes the fastest-evolving taxon of the ingroup and tests for lineage rate homogeneity until all lineages have uniform evolutionary rates. The sequences excluded from the homogeneously evolving taxon subset are flagged as potentially problematic. The software implementation provides the user with the possibility to remove the flagged sequences for generating a new concatenated alignment. We tested LS³ with simulations and two real datasets containing LBA artifacts: a nucleotide dataset regarding the position of Glires within mammals and an amino-acid dataset concerning the position of nematodes within bilaterians. The initially incorrect phylogenies were corrected in all cases upon removing data flagged by LS³. PMID:26912812

A Deep Learning Method to Automatically Identify Reports of Scientifically Rigorous Clinical Research from the Biomedical Literature: Comparative Analytic Study.

PubMed

Del Fiol, Guilherme; Michelson, Matthew; Iorio, Alfonso; Cotoi, Chris; Haynes, R Brian

2018-06-25

A major barrier to the practice of evidence-based medicine is efficiently finding scientifically sound studies on a given clinical topic. To investigate a deep learning approach to retrieve scientifically sound treatment studies from the biomedical literature. We trained a Convolutional Neural Network using a noisy dataset of 403,216 PubMed citations with title and abstract as features. The deep learning model was compared with state-of-the-art search filters, such as PubMed's Clinical Query Broad treatment filter, McMaster's textword search strategy (no Medical Subject Heading, MeSH, terms), and Clinical Query Balanced treatment filter. A previously annotated dataset (Clinical Hedges) was used as the gold standard. The deep learning model obtained significantly lower recall than the Clinical Queries Broad treatment filter (96.9% vs 98.4%; P<.001); and equivalent recall to McMaster's textword search (96.9% vs 97.1%; P=.57) and Clinical Queries Balanced filter (96.9% vs 97.0%; P=.63). Deep learning obtained significantly higher precision than the Clinical Queries Broad filter (34.6% vs 22.4%; P<.001) and McMaster's textword search (34.6% vs 11.8%; P<.001), but was significantly lower than the Clinical Queries Balanced filter (34.6% vs 40.9%; P<.001). Deep learning performed well compared to state-of-the-art search filters, especially when citations were not indexed. Unlike previous machine learning approaches, the proposed deep learning model does not require feature engineering, or time-sensitive or proprietary features, such as MeSH terms and bibliometrics. Deep learning is a promising approach to identifying reports of scientifically rigorous clinical research. Further work is needed to optimize the deep learning model and to assess generalizability to other areas, such as diagnosis, etiology, and prognosis. ©Guilherme Del Fiol, Matthew Michelson, Alfonso Iorio, Chris Cotoi, R Brian Haynes. Originally published in the Journal of Medical Internet Research (http://www.jmir.org), 25.06.2018.

A combination of LongSAGE with Solexa sequencing is well suited to explore the depth and the complexity of transcriptome

PubMed Central

Hanriot, Lucie; Keime, Céline; Gay, Nadine; Faure, Claudine; Dossat, Carole; Wincker, Patrick; Scoté-Blachon, Céline; Peyron, Christelle; Gandrillon, Olivier

2008-01-01

Background "Open" transcriptome analysis methods allow to study gene expression without a priori knowledge of the transcript sequences. As of now, SAGE (Serial Analysis of Gene Expression), LongSAGE and MPSS (Massively Parallel Signature Sequencing) are the mostly used methods for "open" transcriptome analysis. Both LongSAGE and MPSS rely on the isolation of 21 pb tag sequences from each transcript. In contrast to LongSAGE, the high throughput sequencing method used in MPSS enables the rapid sequencing of very large libraries containing several millions of tags, allowing deep transcriptome analysis. However, a bias in the complexity of the transcriptome representation obtained by MPSS was recently uncovered. Results In order to make a deep analysis of mouse hypothalamus transcriptome avoiding the limitation introduced by MPSS, we combined LongSAGE with the Solexa sequencing technology and obtained a library of more than 11 millions of tags. We then compared it to a LongSAGE library of mouse hypothalamus sequenced with the Sanger method. Conclusion We found that Solexa sequencing technology combined with LongSAGE is perfectly suited for deep transcriptome analysis. In contrast to MPSS, it gives a complex representation of transcriptome as reliable as a LongSAGE library sequenced by the Sanger method. PMID:18796152

Identifying Differentially Abundant Metabolic Pathways in Metagenomic Datasets

NASA Astrophysics Data System (ADS)

Liu, Bo; Pop, Mihai

Enabled by rapid advances in sequencing technology, metagenomic studies aim to characterize entire communities of microbes bypassing the need for culturing individual bacterial members. One major goal of such studies is to identify specific functional adaptations of microbial communities to their habitats. Here we describe a powerful analytical method (MetaPath) that can identify differentially abundant pathways in metagenomic data-sets, relying on a combination of metagenomic sequence data and prior metabolic pathway knowledge. We show that MetaPath outperforms other common approaches when evaluated on simulated datasets. We also demonstrate the power of our methods in analyzing two, publicly available, metagenomic datasets: a comparison of the gut microbiome of obese and lean twins; and a comparison of the gut microbiome of infant and adult subjects. We demonstrate that the subpathways identified by our method provide valuable insights into the biological activities of the microbiome.

Simultaneous mutation and copy number variation (CNV) detection by multiplex PCR-based GS-FLX sequencing.

PubMed

Goossens, Dirk; Moens, Lotte N; Nelis, Eva; Lenaerts, An-Sofie; Glassee, Wim; Kalbe, Andreas; Frey, Bruno; Kopal, Guido; De Jonghe, Peter; De Rijk, Peter; Del-Favero, Jurgen

2009-03-01

We evaluated multiplex PCR amplification as a front-end for high-throughput sequencing, to widen the applicability of massive parallel sequencers for the detailed analysis of complex genomes. Using multiplex PCR reactions, we sequenced the complete coding regions of seven genes implicated in peripheral neuropathies in 40 individuals on a GS-FLX genome sequencer (Roche). The resulting dataset showed highly specific and uniform amplification. Comparison of the GS-FLX sequencing data with the dataset generated by Sanger sequencing confirmed the detection of all variants present and proved the sensitivity of the method for mutation detection. In addition, we showed that we could exploit the multiplexed PCR amplicons to determine individual copy number variation (CNV), increasing the spectrum of detected variations to both genetic and genomic variants. We conclude that our straightforward procedure substantially expands the applicability of the massive parallel sequencers for sequencing projects of a moderate number of amplicons (50-500) with typical applications in resequencing exons in positional or functional candidate regions and molecular genetic diagnostics. 2008 Wiley-Liss, Inc.

Breaking barriers and halting rupture: the 2016 Amatrice-Visso-Castelluccio earthquake sequence, central Italy

NASA Astrophysics Data System (ADS)

Gregory, L. C.; Walters, R. J.; Wedmore, L. N. J.; Craig, T. J.; McCaffrey, K. J. W.; Wilkinson, M. W.; Livio, F.; Michetti, A.; Goodall, H.; Li, Z.; Chen, J.; De Martini, P. M.

2017-12-01

In 2016 the Central Italian Apennines was struck by a sequence of normal faulting earthquakes that ruptured in three separate events on the 24th August (Mw 6.2), the 26th Oct (Mw 6.1), and the 30th Oct (Mw 6.6). We reveal the complex nature of the individual events and the time-evolution of the sequence using multiple datasets. We will present an overview of the results from field geology, satellite geodesy, GNSS (including low-cost short baseline installations), and terrestrial laser scanning (TLS). Sequences of earthquakes of mid to high magnitude 6 are common in historical and seismological records in Italy and other similar tectonic settings globally. Multi-fault rupture during these sequences can occur in seconds, as in the M 6.9 1980 Irpinia earthquake, or can span days, months, or years (e.g. the 1703 Norcia-L'Aquila sequence). It is critical to determine why the causative faults in the 2016 sequence did not rupture simultaneously, and how this relates to fault segmentation and structural barriers. This is the first sequence of this kind to be observed using modern geodetic techniques, and only with all of the datasets combined can we begin to understand how and why the sequence evolved in time and space. We show that earthquake rupture both broke through structural barriers that were thought to exist, but was also inhibited by a previously unknown structure. We will also discuss the logistical challenges in generating datasets on the time-evolving sequence, and show how rapid response and international collaboration within the Open EMERGEO Working Group was critical for gaining a complete picture of the ongoing activity.

Genetic Determinants of Drug Resistance in Mycobacterium tuberculosis and Their Diagnostic Value.

PubMed

Farhat, Maha R; Sultana, Razvan; Iartchouk, Oleg; Bozeman, Sam; Galagan, James; Sisk, Peter; Stolte, Christian; Nebenzahl-Guimaraes, Hanna; Jacobson, Karen; Sloutsky, Alexander; Kaur, Devinder; Posey, James; Kreiswirth, Barry N; Kurepina, Natalia; Rigouts, Leen; Streicher, Elizabeth M; Victor, Tommie C; Warren, Robin M; van Soolingen, Dick; Murray, Megan

2016-09-01

The development of molecular diagnostics that detect both the presence of Mycobacterium tuberculosis in clinical samples and drug resistance-conferring mutations promises to revolutionize patient care and interrupt transmission by ensuring early diagnosis. However, these tools require the identification of genetic determinants of resistance to the full range of antituberculosis drugs. To determine the optimal molecular approach needed, we sought to create a comprehensive catalog of resistance mutations and assess their sensitivity and specificity in diagnosing drug resistance. We developed and validated molecular inversion probes for DNA capture and deep sequencing of 28 drug-resistance loci in M. tuberculosis. We used the probes for targeted sequencing of a geographically diverse set of 1,397 clinical M. tuberculosis isolates with known drug resistance phenotypes. We identified a minimal set of mutations to predict resistance to first- and second-line antituberculosis drugs and validated our predictions in an independent dataset. We constructed and piloted a web-based database that provides public access to the sequence data and prediction tool. The predicted resistance to rifampicin and isoniazid exceeded 90% sensitivity and specificity but was lower for other drugs. The number of mutations needed to diagnose resistance is large, and for the 13 drugs studied it was 238 across 18 genetic loci. These data suggest that a comprehensive M. tuberculosis drug resistance diagnostic will need to allow for a high dimension of mutation detection. They also support the hypothesis that currently unknown genetic determinants, potentially discoverable by whole-genome sequencing, encode resistance to second-line tuberculosis drugs.

Genetic Determinants of Drug Resistance in Mycobacterium tuberculosis and Their Diagnostic Value

PubMed Central

Sultana, Razvan; Iartchouk, Oleg; Bozeman, Sam; Galagan, James; Sisk, Peter; Stolte, Christian; Nebenzahl-Guimaraes, Hanna; Jacobson, Karen; Sloutsky, Alexander; Kaur, Devinder; Posey, James; Kreiswirth, Barry N.; Kurepina, Natalia; Rigouts, Leen; Streicher, Elizabeth M.; Victor, Tommie C.; Warren, Robin M.; van Soolingen, Dick; Murray, Megan

2016-01-01

Rationale: The development of molecular diagnostics that detect both the presence of Mycobacterium tuberculosis in clinical samples and drug resistance–conferring mutations promises to revolutionize patient care and interrupt transmission by ensuring early diagnosis. However, these tools require the identification of genetic determinants of resistance to the full range of antituberculosis drugs. Objectives: To determine the optimal molecular approach needed, we sought to create a comprehensive catalog of resistance mutations and assess their sensitivity and specificity in diagnosing drug resistance. Methods: We developed and validated molecular inversion probes for DNA capture and deep sequencing of 28 drug-resistance loci in M. tuberculosis. We used the probes for targeted sequencing of a geographically diverse set of 1,397 clinical M. tuberculosis isolates with known drug resistance phenotypes. We identified a minimal set of mutations to predict resistance to first- and second-line antituberculosis drugs and validated our predictions in an independent dataset. We constructed and piloted a web-based database that provides public access to the sequence data and prediction tool. Measurements and Main Results: The predicted resistance to rifampicin and isoniazid exceeded 90% sensitivity and specificity but was lower for other drugs. The number of mutations needed to diagnose resistance is large, and for the 13 drugs studied it was 238 across 18 genetic loci. Conclusions: These data suggest that a comprehensive M. tuberculosis drug resistance diagnostic will need to allow for a high dimension of mutation detection. They also support the hypothesis that currently unknown genetic determinants, potentially discoverable by whole-genome sequencing, encode resistance to second-line tuberculosis drugs. PMID:26910495

De novo transcriptome assembly databases for the butterfly orchid Phalaenopsis equestris

PubMed Central

Niu, Shan-Ce; Xu, Qing; Zhang, Guo-Qiang; Zhang, Yong-Qiang; Tsai, Wen-Chieh; Hsu, Jui-Ling; Liang, Chieh-Kai; Luo, Yi-Bo; Liu, Zhong-Jian

2016-01-01

Orchids are renowned for their spectacular flowers and ecological adaptations. After the sequencing of the genome of the tropical epiphytic orchid Phalaenopsis equestris, we combined Illumina HiSeq2000 for RNA-Seq and Trinity for de novo assembly to characterize the transcriptomes for 11 diverse P. equestris tissues representing the root, stem, leaf, flower buds, column, lip, petal, sepal and three developmental stages of seeds. Our aims were to contribute to a better understanding of the molecular mechanisms driving the analysed tissue characteristics and to enrich the available data for P. equestris. Here, we present three databases. The first dataset is the RNA-Seq raw reads, which can be used to execute new experiments with different analysis approaches. The other two datasets allow different types of searches for candidate homologues. The second dataset includes the sets of assembled unigenes and predicted coding sequences and proteins, enabling a sequence-based search. The third dataset consists of the annotation results of the aligned unigenes versus the Nonredundant (Nr) protein database, Kyoto Encyclopaedia of Genes and Genomes (KEGG) and Clusters of Orthologous Groups (COG) databases with low e-values, enabling a name-based search. PMID:27673730

Defining the natural fracture network in a shale gas play and its cover succession: The case of the Utica Shale in eastern Canada

NASA Astrophysics Data System (ADS)

Ladevèze, P.; Séjourné, S.; Rivard, C.; Lavoie, D.; Lefebvre, R.; Rouleau, A.

2018-03-01

In the St. Lawrence sedimentary platform (eastern Canada), very little data are available between shallow fresh water aquifers and deep geological hydrocarbon reservoir units (here referred to as the intermediate zone). Characterization of this intermediate zone is crucial, as the latter controls aquifer vulnerability to operations carried out at depth. In this paper, the natural fracture networks in shallow aquifers and in the Utica shale gas reservoir are documented in an attempt to indirectly characterize the intermediate zone. This study used structural data from outcrops, shallow observation well logs and deep shale gas well logs to propose a conceptual model of the natural fracture network. Shallow and deep fractures were categorized into three sets of steeply-dipping fractures and into a set of bedding-parallel fractures. Some lithological and structural controls on fracture distribution were identified. The regional geologic history and similarities between the shallow and deep fracture datasets allowed the extrapolation of the fracture network characterization to the intermediate zone. This study thus highlights the benefits of using both datasets simultaneously, while they are generally interpreted separately. Recommendations are also proposed for future environmental assessment studies in which the existence of preferential flow pathways and potential upward fluid migration toward shallow aquifers need to be identified.

Fast tandem mass spectra-based protein identification regardless of the number of spectra or potential modifications examined.

PubMed

Falkner, Jayson; Andrews, Philip

2005-05-15

Comparing tandem mass spectra (MSMS) against a known dataset of protein sequences is a common method for identifying unknown proteins; however, the processing of MSMS by current software often limits certain applications, including comprehensive coverage of post-translational modifications, non-specific searches and real-time searches to allow result-dependent instrument control. This problem deserves attention as new mass spectrometers provide the ability for higher throughput and as known protein datasets rapidly grow in size. New software algorithms need to be devised in order to address the performance issues of conventional MSMS protein dataset-based protein identification. This paper describes a novel algorithm based on converting a collection of monoisotopic, centroided spectra to a new data structure, named 'peptide finite state machine' (PFSM), which may be used to rapidly search a known dataset of protein sequences, regardless of the number of spectra searched or the number of potential modifications examined. The algorithm is verified using a set of commercially available tryptic digest protein standards analyzed using an ABI 4700 MALDI TOFTOF mass spectrometer, and a free, open source PFSM implementation. It is illustrated that a PFSM can accurately search large collections of spectra against large datasets of protein sequences (e.g. NCBI nr) using a regular desktop PC; however, this paper only details the method for identifying peptide and subsequently protein candidates from a dataset of known protein sequences. The concept of using a PFSM as a peptide pre-screening technique for MSMS-based search engines is validated by using PFSM with Mascot and XTandem. Complete source code, documentation and examples for the reference PFSM implementation are freely available at the Proteome Commons, http://www.proteomecommons.org and source code may be used both commercially and non-commercially as long as the original authors are credited for their work.

Deep learning and texture-based semantic label fusion for brain tumor segmentation

NASA Astrophysics Data System (ADS)

Vidyaratne, L.; Alam, M.; Shboul, Z.; Iftekharuddin, K. M.

2018-02-01

Brain tumor segmentation is a fundamental step in surgical treatment and therapy. Many hand-crafted and learning based methods have been proposed for automatic brain tumor segmentation from MRI. Studies have shown that these approaches have their inherent advantages and limitations. This work proposes a semantic label fusion algorithm by combining two representative state-of-the-art segmentation algorithms: texture based hand-crafted, and deep learning based methods to obtain robust tumor segmentation. We evaluate the proposed method using publicly available BRATS 2017 brain tumor segmentation challenge dataset. The results show that the proposed method offers improved segmentation by alleviating inherent weaknesses: extensive false positives in texture based method, and the false tumor tissue classification problem in deep learning method, respectively. Furthermore, we investigate the effect of patient's gender on the segmentation performance using a subset of validation dataset. Note the substantial improvement in brain tumor segmentation performance proposed in this work has recently enabled us to secure the first place by our group in overall patient survival prediction task at the BRATS 2017 challenge.

Extension of the energy-to-moment parameter Θ to intermediate and deep earthquakes

NASA Astrophysics Data System (ADS)

Saloor, Nooshin; Okal, Emile A.

2018-01-01

We extend to intermediate and deep earthquakes the slowness parameter Θ originally introduced by Newman and Okal (1998). Because of the increasing time lag with depth between the phases P, pP and sP, and of variations in anelastic attenuation parameters t∗ , we define four depth bins featuring slightly different algorithms for the computation of Θ . We apply this methodology to a global dataset of 598 intermediate and deep earthquakes with moments greater than 1025 dyn∗cm. We find a slight increase with depth in average values of Θ (from -4.81 between 80 and 135 km to -4.48 between 450 and 700 km), which however all have intersecting one- σ bands. With widths ranging from 0.26 to 0.31 logarithmic units, these are narrower than their counterpart for a reference dataset of 146 shallow earthquakes (σ = 0.55). Similarly, we find no correlation between values of Θ and focal geometry. These results point to stress conditions within the seismogenic zones inside the Wadati-Benioff slabs more homogeneous than those prevailing at the shallow contacts between tectonic plates.

Deep Learning and Texture-Based Semantic Label Fusion for Brain Tumor Segmentation.

PubMed

Vidyaratne, L; Alam, M; Shboul, Z; Iftekharuddin, K M

2018-01-01

Brain tumor segmentation is a fundamental step in surgical treatment and therapy. Many hand-crafted and learning based methods have been proposed for automatic brain tumor segmentation from MRI. Studies have shown that these approaches have their inherent advantages and limitations. This work proposes a semantic label fusion algorithm by combining two representative state-of-the-art segmentation algorithms: texture based hand-crafted, and deep learning based methods to obtain robust tumor segmentation. We evaluate the proposed method using publicly available BRATS 2017 brain tumor segmentation challenge dataset. The results show that the proposed method offers improved segmentation by alleviating inherent weaknesses: extensive false positives in texture based method, and the false tumor tissue classification problem in deep learning method, respectively. Furthermore, we investigate the effect of patient's gender on the segmentation performance using a subset of validation dataset. Note the substantial improvement in brain tumor segmentation performance proposed in this work has recently enabled us to secure the first place by our group in overall patient survival prediction task at the BRATS 2017 challenge.

Robust visual tracking via multiscale deep sparse networks

NASA Astrophysics Data System (ADS)

Wang, Xin; Hou, Zhiqiang; Yu, Wangsheng; Xue, Yang; Jin, Zefenfen; Dai, Bo

2017-04-01

In visual tracking, deep learning with offline pretraining can extract more intrinsic and robust features. It has significant success solving the tracking drift in a complicated environment. However, offline pretraining requires numerous auxiliary training datasets and is considerably time-consuming for tracking tasks. To solve these problems, a multiscale sparse networks-based tracker (MSNT) under the particle filter framework is proposed. Based on the stacked sparse autoencoders and rectifier linear unit, the tracker has a flexible and adjustable architecture without the offline pretraining process and exploits the robust and powerful features effectively only through online training of limited labeled data. Meanwhile, the tracker builds four deep sparse networks of different scales, according to the target's profile type. During tracking, the tracker selects the matched tracking network adaptively in accordance with the initial target's profile type. It preserves the inherent structural information more efficiently than the single-scale networks. Additionally, a corresponding update strategy is proposed to improve the robustness of the tracker. Extensive experimental results on a large scale benchmark dataset show that the proposed method performs favorably against state-of-the-art methods in challenging environments.

Analyzing the relationship between sequence divergence and nodal support using Bayesian phylogenetic analyses.

PubMed

Makowsky, Robert; Cox, Christian L; Roelke, Corey; Chippindale, Paul T

2010-11-01

Determining the appropriate gene for phylogeny reconstruction can be a difficult process. Rapidly evolving genes tend to resolve recent relationships, but suffer from alignment issues and increased homoplasy among distantly related species. Conversely, slowly evolving genes generally perform best for deeper relationships, but lack sufficient variation to resolve recent relationships. We determine the relationship between sequence divergence and Bayesian phylogenetic reconstruction ability using both natural and simulated datasets. The natural data are based on 28 well-supported relationships within the subphylum Vertebrata. Sequences of 12 genes were acquired and Bayesian analyses were used to determine phylogenetic support for correct relationships. Simulated datasets were designed to determine whether an optimal range of sequence divergence exists across extreme phylogenetic conditions. Across all genes we found that an optimal range of divergence for resolving the correct relationships does exist, although this level of divergence expectedly depends on the distance metric. Simulated datasets show that an optimal range of sequence divergence exists across diverse topologies and models of evolution. We determine that a simple to measure property of genetic sequences (genetic distance) is related to phylogenic reconstruction ability in Bayesian analyses. This information should be useful for selecting the most informative gene to resolve any relationships, especially those that are difficult to resolve, as well as minimizing both cost and confounding information during project design. Copyright © 2010. Published by Elsevier Inc.

A user's guide to quantitative and comparative analysis of metagenomic datasets.

PubMed

Luo, Chengwei; Rodriguez-R, Luis M; Konstantinidis, Konstantinos T

2013-01-01

Metagenomics has revolutionized microbiological studies during the past decade and provided new insights into the diversity, dynamics, and metabolic potential of natural microbial communities. However, metagenomics still represents a field in development, and standardized tools and approaches to handle and compare metagenomes have not been established yet. An important reason accounting for the latter is the continuous changes in the type of sequencing data available, for example, long versus short sequencing reads. Here, we provide a guide to bioinformatic pipelines developed to accomplish the following tasks, focusing primarily on those developed by our team: (i) assemble a metagenomic dataset; (ii) determine the level of sequence coverage obtained and the amount of sequencing required to obtain complete coverage; (iii) identify the taxonomic affiliation of a metagenomic read or assembled contig; and (iv) determine differentially abundant genes, pathways, and species between different datasets. Most of these pipelines do not depend on the type of sequences available or can be easily adjusted to fit different types of sequences, and are freely available (for instance, through our lab Web site: http://www.enve-omics.gatech.edu/). The limitations of current approaches, as well as the computational aspects that can be further improved, will also be briefly discussed. The work presented here provides practical guidelines on how to perform metagenomic analysis of microbial communities characterized by varied levels of diversity and establishes approaches to handle the resulting data, independent of the sequencing platform employed. © 2013 Elsevier Inc. All rights reserved.

An improved filtering algorithm for big read datasets and its application to single-cell assembly.

PubMed

Wedemeyer, Axel; Kliemann, Lasse; Srivastav, Anand; Schielke, Christian; Reusch, Thorsten B; Rosenstiel, Philip

2017-07-03

For single-cell or metagenomic sequencing projects, it is necessary to sequence with a very high mean coverage in order to make sure that all parts of the sample DNA get covered by the reads produced. This leads to huge datasets with lots of redundant data. A filtering of this data prior to assembly is advisable. Brown et al. (2012) presented the algorithm Diginorm for this purpose, which filters reads based on the abundance of their k-mers. We present Bignorm, a faster and quality-conscious read filtering algorithm. An important new algorithmic feature is the use of phred quality scores together with a detailed analysis of the k-mer counts to decide which reads to keep. We qualify and recommend parameters for our new read filtering algorithm. Guided by these parameters, we remove in terms of median 97.15% of the reads while keeping the mean phred score of the filtered dataset high. Using the SDAdes assembler, we produce assemblies of high quality from these filtered datasets in a fraction of the time needed for an assembly from the datasets filtered with Diginorm. We conclude that read filtering is a practical and efficient method for reducing read data and for speeding up the assembly process. This applies not only for single cell assembly, as shown in this paper, but also to other projects with high mean coverage datasets like metagenomic sequencing projects. Our Bignorm algorithm allows assemblies of competitive quality in comparison to Diginorm, while being much faster. Bignorm is available for download at https://git.informatik.uni-kiel.de/axw/Bignorm .

Deep Impact Sequence Planning Using Multi-Mission Adaptable Planning Tools With Integrated Spacecraft Models

NASA Technical Reports Server (NTRS)

Wissler, Steven S.; Maldague, Pierre; Rocca, Jennifer; Seybold, Calina

2006-01-01

The Deep Impact mission was ambitious and challenging. JPL's well proven, easily adaptable multi-mission sequence planning tools combined with integrated spacecraft subsystem models enabled a small operations team to develop, validate, and execute extremely complex sequence-based activities within very short development times. This paper focuses on the core planning tool used in the mission, APGEN. It shows how the multi-mission design and adaptability of APGEN made it possible to model spacecraft subsystems as well as ground assets throughout the lifecycle of the Deep Impact project, starting with models of initial, high-level mission objectives, and culminating in detailed predictions of spacecraft behavior during mission-critical activities.

Transcriptome Sequences Resolve Deep Relationships of the Grape Family

PubMed Central

Wen, Jun; Xiong, Zhiqiang; Nie, Ze-Long; Mao, Likai; Zhu, Yabing; Kan, Xian-Zhao; Ickert-Bond, Stefanie M.; Gerrath, Jean; Zimmer, Elizabeth A.; Fang, Xiao-Dong

2013-01-01

Previous phylogenetic studies of the grape family (Vitaceae) yielded poorly resolved deep relationships, thus impeding our understanding of the evolution of the family. Next-generation sequencing now offers access to protein coding sequences very easily, quickly and cost-effectively. To improve upon earlier work, we extracted 417 orthologous single-copy nuclear genes from the transcriptomes of 15 species of the Vitaceae, covering its phylogenetic diversity. The resulting transcriptome phylogeny provides robust support for the deep relationships, showing the phylogenetic utility of transcriptome data for plants over a time scale at least since the mid-Cretaceous. The pros and cons of transcriptome data for phylogenetic inference in plants are also evaluated. PMID:24069307

PARRoT- a homology-based strategy to quantify and compare RNA-sequencing from non-model organisms.

PubMed

Gan, Ruei-Chi; Chen, Ting-Wen; Wu, Timothy H; Huang, Po-Jung; Lee, Chi-Ching; Yeh, Yuan-Ming; Chiu, Cheng-Hsun; Huang, Hsien-Da; Tang, Petrus

2016-12-22

Next-generation sequencing promises the de novo genomic and transcriptomic analysis of samples of interests. However, there are only a few organisms having reference genomic sequences and even fewer having well-defined or curated annotations. For transcriptome studies focusing on organisms lacking proper reference genomes, the common strategy is de novo assembly followed by functional annotation. However, things become even more complicated when multiple transcriptomes are compared. Here, we propose a new analysis strategy and quantification methods for quantifying expression level which not only generate a virtual reference from sequencing data, but also provide comparisons between transcriptomes. First, all reads from the transcriptome datasets are pooled together for de novo assembly. The assembled contigs are searched against NCBI NR databases to find potential homolog sequences. Based on the searched result, a set of virtual transcripts are generated and served as a reference transcriptome. By using the same reference, normalized quantification values including RC (read counts), eRPKM (estimated RPKM) and eTPM (estimated TPM) can be obtained that are comparable across transcriptome datasets. In order to demonstrate the feasibility of our strategy, we implement it in the web service PARRoT. PARRoT stands for Pipeline for Analyzing RNA Reads of Transcriptomes. It analyzes gene expression profiles for two transcriptome sequencing datasets. For better understanding of the biological meaning from the comparison among transcriptomes, PARRoT further provides linkage between these virtual transcripts and their potential function through showing best hits in SwissProt, NR database, assigning GO terms. Our demo datasets showed that PARRoT can analyze two paired-end transcriptomic datasets of approximately 100 million reads within just three hours. In this study, we proposed and implemented a strategy to analyze transcriptomes from non-reference organisms which offers the opportunity to quantify and compare transcriptome profiles through a homolog based virtual transcriptome reference. By using the homolog based reference, our strategy effectively avoids the problems that may cause from inconsistencies among transcriptomes. This strategy will shed lights on the field of comparative genomics for non-model organism. We have implemented PARRoT as a web service which is freely available at http://parrot.cgu.edu.tw .

Deep multi-scale convolutional neural network for hyperspectral image classification

NASA Astrophysics Data System (ADS)

Zhang, Feng-zhe; Yang, Xia

2018-04-01

In this paper, we proposed a multi-scale convolutional neural network for hyperspectral image classification task. Firstly, compared with conventional convolution, we utilize multi-scale convolutions, which possess larger respective fields, to extract spectral features of hyperspectral image. We design a deep neural network with a multi-scale convolution layer which contains 3 different convolution kernel sizes. Secondly, to avoid overfitting of deep neural network, dropout is utilized, which randomly sleeps neurons, contributing to improve the classification accuracy a bit. In addition, new skills like ReLU in deep learning is utilized in this paper. We conduct experiments on University of Pavia and Salinas datasets, and obtained better classification accuracy compared with other methods.

Inadequate Reference Datasets Biased toward Short Non-epitopes Confound B-cell Epitope Prediction*

PubMed Central

Rahman, Kh. Shamsur; Chowdhury, Erfan Ullah; Sachse, Konrad; Kaltenboeck, Bernhard

2016-01-01

X-ray crystallography has shown that an antibody paratope typically binds 15–22 amino acids (aa) of an epitope, of which 2–5 randomly distributed amino acids contribute most of the binding energy. In contrast, researchers typically choose for B-cell epitope mapping short peptide antigens in antibody binding assays. Furthermore, short 6–11-aa epitopes, and in particular non-epitopes, are over-represented in published B-cell epitope datasets that are commonly used for development of B-cell epitope prediction approaches from protein antigen sequences. We hypothesized that such suboptimal length peptides result in weak antibody binding and cause false-negative results. We tested the influence of peptide antigen length on antibody binding by analyzing data on more than 900 peptides used for B-cell epitope mapping of immunodominant proteins of Chlamydia spp. We demonstrate that short 7–12-aa peptides of B-cell epitopes bind antibodies poorly; thus, epitope mapping with short peptide antigens falsely classifies many B-cell epitopes as non-epitopes. We also show in published datasets of confirmed epitopes and non-epitopes a direct correlation between length of peptide antigens and antibody binding. Elimination of short, ≤11-aa epitope/non-epitope sequences improved datasets for evaluation of in silico B-cell epitope prediction. Achieving up to 86% accuracy, protein disorder tendency is the best indicator of B-cell epitope regions for chlamydial and published datasets. For B-cell epitope prediction, the most effective approach is plotting disorder of protein sequences with the IUPred-L scale, followed by antibody reactivity testing of 16–30-aa peptides from peak regions. This strategy overcomes the well known inaccuracy of in silico B-cell epitope prediction from primary protein sequences. PMID:27189949

Deep Learning and Its Applications in Biomedicine.

PubMed

Cao, Chensi; Liu, Feng; Tan, Hai; Song, Deshou; Shu, Wenjie; Li, Weizhong; Zhou, Yiming; Bo, Xiaochen; Xie, Zhi

2018-02-01

Advances in biological and medical technologies have been providing us explosive volumes of biological and physiological data, such as medical images, electroencephalography, genomic and protein sequences. Learning from these data facilitates the understanding of human health and disease. Developed from artificial neural networks, deep learning-based algorithms show great promise in extracting features and learning patterns from complex data. The aim of this paper is to provide an overview of deep learning techniques and some of the state-of-the-art applications in the biomedical field. We first introduce the development of artificial neural network and deep learning. We then describe two main components of deep learning, i.e., deep learning architectures and model optimization. Subsequently, some examples are demonstrated for deep learning applications, including medical image classification, genomic sequence analysis, as well as protein structure classification and prediction. Finally, we offer our perspectives for the future directions in the field of deep learning. Copyright © 2018. Production and hosting by Elsevier B.V.

Analyses of mitochondrial amino acid sequence datasets support the proposal that specimens of Hypodontus macropi from three species of macropodid hosts represent distinct species

PubMed Central

2013-01-01

Background Hypodontus macropi is a common intestinal nematode of a range of kangaroos and wallabies (macropodid marsupials). Based on previous multilocus enzyme electrophoresis (MEE) and nuclear ribosomal DNA sequence data sets, H. macropi has been proposed to be complex of species. To test this proposal using independent molecular data, we sequenced the whole mitochondrial (mt) genomes of individuals of H. macropi from three different species of hosts (Macropus robustus robustus, Thylogale billardierii and Macropus [Wallabia] bicolor) as well as that of Macropicola ocydromi (a related nematode), and undertook a comparative analysis of the amino acid sequence datasets derived from these genomes. Results The mt genomes sequenced by next-generation (454) technology from H. macropi from the three host species varied from 13,634 bp to 13,699 bp in size. Pairwise comparisons of the amino acid sequences predicted from these three mt genomes revealed differences of 5.8% to 18%. Phylogenetic analysis of the amino acid sequence data sets using Bayesian Inference (BI) showed that H. macropi from the three different host species formed distinct, well-supported clades. In addition, sliding window analysis of the mt genomes defined variable regions for future population genetic studies of H. macropi in different macropodid hosts and geographical regions around Australia. Conclusions The present analyses of inferred mt protein sequence datasets clearly supported the hypothesis that H. macropi from M. robustus robustus, M. bicolor and T. billardierii represent distinct species. PMID:24261823

HMMER Cut-off Threshold Tool (HMMERCTTER): Supervised classification of superfamily protein sequences with a reliable cut-off threshold.

PubMed

Pagnuco, Inti Anabela; Revuelta, María Victoria; Bondino, Hernán Gabriel; Brun, Marcel; Ten Have, Arjen

2018-01-01

Protein superfamilies can be divided into subfamilies of proteins with different functional characteristics. Their sequences can be classified hierarchically, which is part of sequence function assignation. Typically, there are no clear subfamily hallmarks that would allow pattern-based function assignation by which this task is mostly achieved based on the similarity principle. This is hampered by the lack of a score cut-off that is both sensitive and specific. HMMER Cut-off Threshold Tool (HMMERCTTER) adds a reliable cut-off threshold to the popular HMMER. Using a high quality superfamily phylogeny, it clusters a set of training sequences such that the cluster-specific HMMER profiles show cluster or subfamily member detection with 100% precision and recall (P&R), thereby generating a specific threshold as inclusion cut-off. Profiles and thresholds are then used as classifiers to screen a target dataset. Iterative inclusion of novel sequences to groups and the corresponding HMMER profiles results in high sensitivity while specificity is maintained by imposing 100% P&R self detection. In three presented case studies of protein superfamilies, classification of large datasets with 100% precision was achieved with over 95% recall. Limits and caveats are presented and explained. HMMERCTTER is a promising protein superfamily sequence classifier provided high quality training datasets are used. It provides a decision support system that aids in the difficult task of sequence function assignation in the twilight zone of sequence similarity. All relevant data and source codes are available from the Github repository at the following URL: https://github.com/BBCMdP/HMMERCTTER.

HMMER Cut-off Threshold Tool (HMMERCTTER): Supervised classification of superfamily protein sequences with a reliable cut-off threshold

PubMed Central

Pagnuco, Inti Anabela; Revuelta, María Victoria; Bondino, Hernán Gabriel; Brun, Marcel

2018-01-01

Background Protein superfamilies can be divided into subfamilies of proteins with different functional characteristics. Their sequences can be classified hierarchically, which is part of sequence function assignation. Typically, there are no clear subfamily hallmarks that would allow pattern-based function assignation by which this task is mostly achieved based on the similarity principle. This is hampered by the lack of a score cut-off that is both sensitive and specific. Results HMMER Cut-off Threshold Tool (HMMERCTTER) adds a reliable cut-off threshold to the popular HMMER. Using a high quality superfamily phylogeny, it clusters a set of training sequences such that the cluster-specific HMMER profiles show cluster or subfamily member detection with 100% precision and recall (P&R), thereby generating a specific threshold as inclusion cut-off. Profiles and thresholds are then used as classifiers to screen a target dataset. Iterative inclusion of novel sequences to groups and the corresponding HMMER profiles results in high sensitivity while specificity is maintained by imposing 100% P&R self detection. In three presented case studies of protein superfamilies, classification of large datasets with 100% precision was achieved with over 95% recall. Limits and caveats are presented and explained. Conclusions HMMERCTTER is a promising protein superfamily sequence classifier provided high quality training datasets are used. It provides a decision support system that aids in the difficult task of sequence function assignation in the twilight zone of sequence similarity. All relevant data and source codes are available from the Github repository at the following URL: https://github.com/BBCMdP/HMMERCTTER. PMID:29579071

A novel method based on new adaptive LVQ neural network for predicting protein-protein interactions from protein sequences.

PubMed

Yousef, Abdulaziz; Moghadam Charkari, Nasrollah

2013-11-07

Protein-Protein interaction (PPI) is one of the most important data in understanding the cellular processes. Many interesting methods have been proposed in order to predict PPIs. However, the methods which are based on the sequence of proteins as a prior knowledge are more universal. In this paper, a sequence-based, fast, and adaptive PPI prediction method is introduced to assign two proteins to an interaction class (yes, no). First, in order to improve the presentation of the sequences, twelve physicochemical properties of amino acid have been used by different representation methods to transform the sequence of protein pairs into different feature vectors. Then, for speeding up the learning process and reducing the effect of noise PPI data, principal component analysis (PCA) is carried out as a proper feature extraction algorithm. Finally, a new and adaptive Learning Vector Quantization (LVQ) predictor is designed to deal with different models of datasets that are classified into balanced and imbalanced datasets. The accuracy of 93.88%, 90.03%, and 89.72% has been found on S. cerevisiae, H. pylori, and independent datasets, respectively. The results of various experiments indicate the efficiency and validity of the method. © 2013 Published by Elsevier Ltd.

Emergent HIV-1 Drug Resistance Mutations Were Not Present at Low-Frequency at Baseline in Non-Nucleoside Reverse Transcriptase Inhibitor-Treated Subjects in the STaR Study

PubMed Central

Porter, Danielle P.; Daeumer, Martin; Thielen, Alexander; Chang, Silvia; Martin, Ross; Cohen, Cal; Miller, Michael D.; White, Kirsten L.

2015-01-01

At Week 96 of the Single-Tablet Regimen (STaR) study, more treatment-naïve subjects that received rilpivirine/emtricitabine/tenofovir DF (RPV/FTC/TDF) developed resistance mutations compared to those treated with efavirenz (EFV)/FTC/TDF by population sequencing. Furthermore, more RPV/FTC/TDF-treated subjects with baseline HIV-1 RNA >100,000 copies/mL developed resistance compared to subjects with baseline HIV-1 RNA ≤100,000 copies/mL. Here, deep sequencing was utilized to assess the presence of pre-existing low-frequency variants in subjects with and without resistance development in the STaR study. Deep sequencing (Illumina MiSeq) was performed on baseline and virologic failure samples for all subjects analyzed for resistance by population sequencing during the clinical study (n = 33), as well as baseline samples from control subjects with virologic response (n = 118). Primary NRTI or NNRTI drug resistance mutations present at low frequency (≥2% to 20%) were detected in 6.6% of baseline samples by deep sequencing, all of which occurred in control subjects. Deep sequencing results were generally consistent with population sequencing but detected additional primary NNRTI and NRTI resistance mutations at virologic failure in seven samples. HIV-1 drug resistance mutations emerging while on RPV/FTC/TDF or EFV/FTC/TDF treatment were not present at low frequency at baseline in the STaR study. PMID:26690199

Emergent HIV-1 Drug Resistance Mutations Were Not Present at Low-Frequency at Baseline in Non-Nucleoside Reverse Transcriptase Inhibitor-Treated Subjects in the STaR Study.

PubMed

Porter, Danielle P; Daeumer, Martin; Thielen, Alexander; Chang, Silvia; Martin, Ross; Cohen, Cal; Miller, Michael D; White, Kirsten L

2015-12-07

At Week 96 of the Single-Tablet Regimen (STaR) study, more treatment-naïve subjects that received rilpivirine/emtricitabine/tenofovir DF (RPV/FTC/TDF) developed resistance mutations compared to those treated with efavirenz (EFV)/FTC/TDF by population sequencing. Furthermore, more RPV/FTC/TDF-treated subjects with baseline HIV-1 RNA >100,000 copies/mL developed resistance compared to subjects with baseline HIV-1 RNA ≤100,000 copies/mL. Here, deep sequencing was utilized to assess the presence of pre-existing low-frequency variants in subjects with and without resistance development in the STaR study. Deep sequencing (Illumina MiSeq) was performed on baseline and virologic failure samples for all subjects analyzed for resistance by population sequencing during the clinical study (n = 33), as well as baseline samples from control subjects with virologic response (n = 118). Primary NRTI or NNRTI drug resistance mutations present at low frequency (≥2% to 20%) were detected in 6.6% of baseline samples by deep sequencing, all of which occurred in control subjects. Deep sequencing results were generally consistent with population sequencing but detected additional primary NNRTI and NRTI resistance mutations at virologic failure in seven samples. HIV-1 drug resistance mutations emerging while on RPV/FTC/TDF or EFV/FTC/TDF treatment were not present at low frequency at baseline in the STaR study.

VirusDetect: An automated pipeline for efficient virus discovery using deep sequencing of small RNAs

USDA-ARS?s Scientific Manuscript database

Accurate detection of viruses in plants and animals is critical for agriculture production and human health. Deep sequencing and assembly of virus-derived siRNAs has proven to be a highly efficient approach for virus discovery. However, to date no computational tools specifically designed for both k...

Natural Variation in Brachypodium disctachyon: Deep Sequencing of Highly Diverse Natural Accessions (2013 DOE JGI Genomics of Energy and Environment 8th Annual User Meeting)

DOE Office of Scientific and Technical Information (OSTI.GOV)

Gordon, Sean

2013-03-01

Sean Gordon of the USDA on Natural variation in Brachypodium disctachyon: Deep Sequencing of Highly Diverse Natural Accessions at the 8th Annual Genomics of Energy Environment Meeting on March 27, 2013 in Walnut Creek, CA.

Microbial Diversity in Deep-sea Methane Seep Sediments Presented by SSU rRNA Gene Tag Sequencing

PubMed Central

Nunoura, Takuro; Takaki, Yoshihiro; Kazama, Hiromi; Hirai, Miho; Ashi, Juichiro; Imachi, Hiroyuki; Takai, Ken

2012-01-01

Microbial community structures in methane seep sediments in the Nankai Trough were analyzed by tag-sequencing analysis for the small subunit (SSU) rRNA gene using a newly developed primer set. The dominant members of Archaea were Deep-sea Hydrothermal Vent Euryarchaeotic Group 6 (DHVEG 6), Marine Group I (MGI) and Deep Sea Archaeal Group (DSAG), and those in Bacteria were Alpha-, Gamma-, Delta- and Epsilonproteobacteria, Chloroflexi, Bacteroidetes, Planctomycetes and Acidobacteria. Diversity and richness were examined by 8,709 and 7,690 tag-sequences from sediments at 5 and 25 cm below the seafloor (cmbsf), respectively. The estimated diversity and richness in the methane seep sediment are as high as those in soil and deep-sea hydrothermal environments, although the tag-sequences obtained in this study were not sufficient to show whole microbial diversity in this analysis. We also compared the diversity and richness of each taxon/division between the sediments from the two depths, and found that the diversity and richness of some taxa/divisions varied significantly along with the depth. PMID:22510646

Plant Species Identification by Bi-channel Deep Convolutional Networks

NASA Astrophysics Data System (ADS)

He, Guiqing; Xia, Zhaoqiang; Zhang, Qiqi; Zhang, Haixi; Fan, Jianping

2018-04-01

Plant species identification achieves much attention recently as it has potential application in the environmental protection and human life. Although deep learning techniques can be directly applied for plant species identification, it still needs to be designed for this specific task to obtain the state-of-art performance. In this paper, a bi-channel deep learning framework is developed for identifying plant species. In the framework, two different sub-networks are fine-tuned over their pretrained models respectively. And then a stacking layer is used to fuse the output of two different sub-networks. We construct a plant dataset of Orchidaceae family for algorithm evaluation. Our experimental results have demonstrated that our bi-channel deep network can achieve very competitive performance on accuracy rates compared to the existing deep learning algorithm.

Episodic inflation of Akutan volcano, Alaska revealed from GPS and InSAR time series

NASA Astrophysics Data System (ADS)

DeGrandpre, K.; Lu, Z.; Wang, T.

2016-12-01

Akutan volcano is one of the most active volcanoes located long the Aleutian arc. At least 27 eruptions have been noted since 1790 and an intense swarm of volcano-tectonic earthquakes occurred in 1996. Surface deformation after the 1996 earthquake sequence has been studied using GPS and Interferometric Synthetic Aperture Radar (InSAR) separately, yet models created from these datasets require different mechanisms to produce the observed surface deformation: an inflating Mogi source results in the best approximation of displacement observed from GPS data, whereas an opening dyke is the best fit to deformation measured from InSAR. A recent study using seismic data revealed complex magmatic structures beneath the caldera, suggesting that the surface deformation may reflect more complicated mechanisms that cannot be estimated using one type of data alone. Here we integrate the surface deformation measured from GPS and InSAR to better understand the magma plumbing system beneath Akutan volcano. GPS time-series at 12 stations from 2006 to 2016 were analyzed, and two transient episodes of inflation in 2008 and 2014 were detected. These GPS stations are, however, too sparse to reveal the spatial distribution of the surface deformation. In order to better define the spatial extent of this inflation four tracks of Envisat data acquired during 2003-2010 and one track of TerraSAR-X data acquired from 2010 to 2016 were processed to produce high-resolution maps of surface deformation. These deformation maps show a consistently uplifting area on the northwestern flank of the volcano. We inverted for the source parameters required to produce the inflation using GPS, InSAR, and a dataset of GPS and InSAR measurements combined, to find that a deep Mogi source below a shallow dyke fit these datasets best. From the TerraSAR-X data, we were also able to measure the subsidence inside the summit caldera due to fumarole activity to be as high as 10 mm/yr. The complex spatial and temporal deformation patterns observed using GPS and InSAR at Akutan volcano imply that the magma plumbing system beneath the island inflates episodically from both deep and shallow sources of varying geometry which is responsible for the uplift observed in 2008 and 2014, but has yet led to an eruption.

Skeleton-based human action recognition using multiple sequence alignment

NASA Astrophysics Data System (ADS)

Ding, Wenwen; Liu, Kai; Cheng, Fei; Zhang, Jin; Li, YunSong

2015-05-01

Human action recognition and analysis is an active research topic in computer vision for many years. This paper presents a method to represent human actions based on trajectories consisting of 3D joint positions. This method first decompose action into a sequence of meaningful atomic actions (actionlets), and then label actionlets with English alphabets according to the Davies-Bouldin index value. Therefore, an action can be represented using a sequence of actionlet symbols, which will preserve the temporal order of occurrence of each of the actionlets. Finally, we employ sequence comparison to classify multiple actions through using string matching algorithms (Needleman-Wunsch). The effectiveness of the proposed method is evaluated on datasets captured by commodity depth cameras. Experiments of the proposed method on three challenging 3D action datasets show promising results.

Nebula--a web-server for advanced ChIP-seq data analysis.

PubMed

Boeva, Valentina; Lermine, Alban; Barette, Camille; Guillouf, Christel; Barillot, Emmanuel

2012-10-01

ChIP-seq consists of chromatin immunoprecipitation and deep sequencing of the extracted DNA fragments. It is the technique of choice for accurate characterization of the binding sites of transcription factors and other DNA-associated proteins. We present a web service, Nebula, which allows inexperienced users to perform a complete bioinformatics analysis of ChIP-seq data. Nebula was designed for both bioinformaticians and biologists. It is based on the Galaxy open source framework. Galaxy already includes a large number of functionalities for mapping reads and peak calling. We added the following to Galaxy: (i) peak calling with FindPeaks and a module for immunoprecipitation quality control, (ii) de novo motif discovery with ChIPMunk, (iii) calculation of the density and the cumulative distribution of peak locations relative to gene transcription start sites, (iv) annotation of peaks with genomic features and (v) annotation of genes with peak information. Nebula generates the graphs and the enrichment statistics at each step of the process. During Steps 3-5, Nebula optionally repeats the analysis on a control dataset and compares these results with those from the main dataset. Nebula can also incorporate gene expression (or gene modulation) data during these steps. In summary, Nebula is an innovative web service that provides an advanced ChIP-seq analysis pipeline providing ready-to-publish results. Nebula is available at http://nebula.curie.fr/ Supplementary data are available at Bioinformatics online.

Draft Genome Sequence of Pseudomonas oceani DSM 100277T, a Deep-Sea Bacterium.

PubMed

García-Valdés, Elena; Gomila, Margarita; Mulet, Magdalena; Lalucat, Jorge

2018-04-12

Pseudomonas oceani DSM 100277 T was isolated from deep seawater in the Okinawa Trough at 1390 m. P. oceani belongs to the Pseudomonas pertucinogena group. Here, we report the draft genome sequence of P. oceani , which has an estimated size of 4.1 Mb and exhibits 3,790 coding sequences, with a G+C content of 59.94 mol%. Copyright © 2018 García-Valdés et al.

MIPS bacterial genomes functional annotation benchmark dataset.

PubMed

Tetko, Igor V; Brauner, Barbara; Dunger-Kaltenbach, Irmtraud; Frishman, Goar; Montrone, Corinna; Fobo, Gisela; Ruepp, Andreas; Antonov, Alexey V; Surmeli, Dimitrij; Mewes, Hans-Wernen

2005-05-15

Any development of new methods for automatic functional annotation of proteins according to their sequences requires high-quality data (as benchmark) as well as tedious preparatory work to generate sequence parameters required as input data for the machine learning methods. Different program settings and incompatible protocols make a comparison of the analyzed methods difficult. The MIPS Bacterial Functional Annotation Benchmark dataset (MIPS-BFAB) is a new, high-quality resource comprising four bacterial genomes manually annotated according to the MIPS functional catalogue (FunCat). These resources include precalculated sequence parameters, such as sequence similarity scores, InterPro domain composition and other parameters that could be used to develop and benchmark methods for functional annotation of bacterial protein sequences. These data are provided in XML format and can be used by scientists who are not necessarily experts in genome annotation. BFAB is available at http://mips.gsf.de/proj/bfab

Deep Ion Torrent sequencing identifies soil fungal community shifts after frequent prescribed fires in a southeastern US forest ecosystem.

PubMed

Brown, Shawn P; Callaham, Mac A; Oliver, Alena K; Jumpponen, Ari

2013-12-01

Prescribed burning is a common management tool to control fuel loads, ground vegetation, and facilitate desirable game species. We evaluated soil fungal community responses to long-term prescribed fire treatments in a loblolly pine forest on the Piedmont of Georgia and utilized deep Internal Transcribed Spacer Region 1 (ITS1) amplicon sequencing afforded by the recent Ion Torrent Personal Genome Machine (PGM). These deep sequence data (19,000 + reads per sample after subsampling) indicate that frequent fires (3-year fire interval) shift soil fungus communities, whereas infrequent fires (6-year fire interval) permit system resetting to a state similar to that without prescribed fire. Furthermore, in nonmetric multidimensional scaling analyses, primarily ectomycorrhizal taxa were correlated with axes associated with long fire intervals, whereas soil saprobes tended to be correlated with the frequent fire recurrence. We conclude that (1) multiplexed Ion Torrent PGM analyses allow deep cost effective sequencing of fungal communities but may suffer from short read lengths and inconsistent sequence quality adjacent to the sequencing adaptor; (2) frequent prescribed fires elicit a shift in soil fungal communities; and (3) such shifts do not occur when fire intervals are longer. Our results emphasize the general responsiveness of these forests to management, and the importance of fire return intervals in meeting management objectives. © 2013 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved.

Predicting protein-binding regions in RNA using nucleotide profiles and compositions.

PubMed

Choi, Daesik; Park, Byungkyu; Chae, Hanju; Lee, Wook; Han, Kyungsook

2017-03-14

Motivated by the increased amount of data on protein-RNA interactions and the availability of complete genome sequences of several organisms, many computational methods have been proposed to predict binding sites in protein-RNA interactions. However, most computational methods are limited to finding RNA-binding sites in proteins instead of protein-binding sites in RNAs. Predicting protein-binding sites in RNA is more challenging than predicting RNA-binding sites in proteins. Recent computational methods for finding protein-binding sites in RNAs have several drawbacks for practical use. We developed a new support vector machine (SVM) model for predicting protein-binding regions in mRNA sequences. The model uses sequence profiles constructed from log-odds scores of mono- and di-nucleotides and nucleotide compositions. The model was evaluated by standard 10-fold cross validation, leave-one-protein-out (LOPO) cross validation and independent testing. Since actual mRNA sequences have more non-binding regions than protein-binding regions, we tested the model on several datasets with different ratios of protein-binding regions to non-binding regions. The best performance of the model was obtained in a balanced dataset of positive and negative instances. 10-fold cross validation with a balanced dataset achieved a sensitivity of 91.6%, a specificity of 92.4%, an accuracy of 92.0%, a positive predictive value (PPV) of 91.7%, a negative predictive value (NPV) of 92.3% and a Matthews correlation coefficient (MCC) of 0.840. LOPO cross validation showed a lower performance than the 10-fold cross validation, but the performance remains high (87.6% accuracy and 0.752 MCC). In testing the model on independent datasets, it achieved an accuracy of 82.2% and an MCC of 0.656. Testing of our model and other state-of-the-art methods on a same dataset showed that our model is better than the others. Sequence profiles of log-odds scores of mono- and di-nucleotides were much more powerful features than nucleotide compositions in finding protein-binding regions in RNA sequences. But, a slight performance gain was obtained when using the sequence profiles along with nucleotide compositions. These are preliminary results of ongoing research, but demonstrate the potential of our approach as a powerful predictor of protein-binding regions in RNA. The program and supporting data are available at http://bclab.inha.ac.kr/RBPbinding .

Development of a EST dataset and characterization of EST-SSRs in a traditional Chinese medicinal plant, Epimedium sagittatum (Sieb. Et Zucc.) Maxim

PubMed Central

2010-01-01

Background Epimedium sagittatum (Sieb. Et Zucc.) Maxim, a traditional Chinese medicinal plant species, has been used extensively as genuine medicinal materials. Certain Epimedium species are endangered due to commercial overexploition, while sustainable application studies, conservation genetics, systematics, and marker-assisted selection (MAS) of Epimedium is less-studied due to the lack of molecular markers. Here, we report a set of expressed sequence tags (ESTs) and simple sequence repeats (SSRs) identified in these ESTs for E. sagittatum. Results cDNAs of E. sagittatum are sequenced using 454 GS-FLX pyrosequencing technology. The raw reads are cleaned and assembled into a total of 76,459 consensus sequences comprising of 17,231 contigs and 59,228 singlets. About 38.5% (29,466) of the consensus sequences significantly match to the non-redundant protein database (E-value < 1e-10), 22,295 of which are further annotated using Gene Ontology (GO) terms. A total of 2,810 EST-SSRs is identified from the Epimedium EST dataset. Trinucleotide SSR is the dominant repeat type (55.2%) followed by dinucleotide (30.4%), tetranuleotide (7.3%), hexanucleotide (4.9%), and pentanucleotide (2.2%) SSR. The dominant repeat motif is AAG/CTT (23.6%) followed by AG/CT (19.3%), ACC/GGT (11.1%), AT/AT (7.5%), and AAC/GTT (5.9%). Thirty-two SSR-ESTs are randomly selected and primer pairs are synthesized for testing the transferability across 52 Epimedium species. Eighteen primer pairs (85.7%) could be successfully transferred to Epimedium species and sixteen of those show high genetic diversity with 0.35 of observed heterozygosity (Ho) and 0.65 of expected heterozygosity (He) and high number of alleles per locus (11.9). Conclusion A large EST dataset with a total of 76,459 consensus sequences is generated, aiming to provide sequence information for deciphering secondary metabolism, especially for flavonoid pathway in Epimedium. A total of 2,810 EST-SSRs is identified from EST dataset and ~1580 EST-SSR markers are transferable. E. sagittatum EST-SSR transferability to the major Epimedium germplasm is up to 85.7%. Therefore, this EST dataset and EST-SSRs will be a powerful resource for further studies such as taxonomy, molecular breeding, genetics, genomics, and secondary metabolism in Epimedium species. PMID:20141623

RNA-Seq analysis to capture the transcriptome landscape of a single cell

PubMed Central

Tang, Fuchou; Barbacioru, Catalin; Nordman, Ellen; Xu, Nanlan; Bashkirov, Vladimir I; Lao, Kaiqin; Surani, M. Azim

2013-01-01

We describe here a protocol for digital transcriptome analysis in a single mouse blastomere using a deep sequencing approach. An individual blastomere was first isolated and put into lysate buffer by mouth pipette. Reverse transcription was then performed directly on the whole cell lysate. After this, the free primers were removed by Exonuclease I and a poly(A) tail was added to the 3′ end of the first-strand cDNA by Terminal Deoxynucleotidyl Transferase. Then the single cell cDNAs were amplified by 20 plus 9 cycles of PCR. Then 100-200 ng of these amplified cDNAs were used to construct a sequencing library. The sequencing library can be used for deep sequencing using the SOLiD system. Compared with the cDNA microarray technique, our assay can capture up to 75% more genes expressed in early embryos. The protocol can generate deep sequencing libraries within 6 days for 16 single cell samples. PMID:20203668

Deep sequencing reveals double mutations in cis of MPL exon 10 in myeloproliferative neoplasms.

PubMed

Pietra, Daniela; Brisci, Angela; Rumi, Elisa; Boggi, Sabrina; Elena, Chiara; Pietrelli, Alessandro; Bordoni, Roberta; Ferrari, Maurizio; Passamonti, Francesco; De Bellis, Gianluca; Cremonesi, Laura; Cazzola, Mario

2011-04-01

Somatic mutations of MPL exon 10, mainly involving a W515 substitution, have been described in JAK2 (V617F)-negative patients with essential thrombocythemia and primary myelofibrosis. We used direct sequencing and high-resolution melt analysis to identify mutations of MPL exon 10 in 570 patients with myeloproliferative neoplasms, and allele specific PCR and deep sequencing to further characterize a subset of mutated patients. Somatic mutations were detected in 33 of 221 patients (15%) with JAK2 (V617F)-negative essential thrombocythemia or primary myelofibrosis. Only one patient with essential thrombocythemia carried both JAK2 (V617F) and MPL (W515L). High-resolution melt analysis identified abnormal patterns in all the MPL mutated cases, while direct sequencing did not detect the mutant MPL in one fifth of them. In 3 cases carrying double MPL mutations, deep sequencing analysis showed identical load and location in cis of the paired lesions, indicating their simultaneous occurrence on the same chromosome.

Evaluation of privacy in high dynamic range video sequences

NASA Astrophysics Data System (ADS)

Řeřábek, Martin; Yuan, Lin; Krasula, Lukáš; Korshunov, Pavel; Fliegel, Karel; Ebrahimi, Touradj

2014-09-01

The ability of high dynamic range (HDR) to capture details in environments with high contrast has a significant impact on privacy in video surveillance. However, the extent to which HDR imaging affects privacy, when compared to a typical low dynamic range (LDR) imaging, is neither well studied nor well understood. To achieve such an objective, a suitable dataset of images and video sequences is needed. Therefore, we have created a publicly available dataset of HDR video for privacy evaluation PEViD-HDR, which is an HDR extension of an existing Privacy Evaluation Video Dataset (PEViD). PEViD-HDR video dataset can help in the evaluations of privacy protection tools, as well as for showing the importance of HDR imaging in video surveillance applications and its influence on the privacy-intelligibility trade-off. We conducted a preliminary subjective experiment demonstrating the usability of the created dataset for evaluation of privacy issues in video. The results confirm that a tone-mapped HDR video contains more privacy sensitive information and details compared to a typical LDR video.

Deep classification hashing for person re-identification

NASA Astrophysics Data System (ADS)

Wang, Jiabao; Li, Yang; Zhang, Xiancai; Miao, Zhuang; Tao, Gang

2018-04-01

As the development of surveillance in public, person re-identification becomes more and more important. The largescale databases call for efficient computation and storage, hashing technique is one of the most important methods. In this paper, we proposed a new deep classification hashing network by introducing a new binary appropriation layer in the traditional ImageNet pre-trained CNN models. It outputs binary appropriate features, which can be easily quantized into binary hash-codes for hamming similarity comparison. Experiments show that our deep hashing method can outperform the state-of-the-art methods on the public CUHK03 and Market1501 datasets.

Spectra library assisted de novo peptide sequencing for HCD and ETD spectra pairs.

PubMed

Yan, Yan; Zhang, Kaizhong

2016-12-23

De novo peptide sequencing via tandem mass spectrometry (MS/MS) has been developed rapidly in recent years. With the use of spectra pairs from the same peptide under different fragmentation modes, performance of de novo sequencing is greatly improved. Currently, with large amount of spectra sequenced everyday, spectra libraries containing tens of thousands of annotated experimental MS/MS spectra become available. These libraries provide information of the spectra properties, thus have the potential to be used with de novo sequencing to improve its performance. In this study, an improved de novo sequencing method assisted with spectra library is proposed. It uses spectra libraries as training datasets and introduces significant scores of the features used in our previous de novo sequencing method for HCD and ETD spectra pairs. Two pairs of HCD and ETD spectral datasets were used to test the performance of the proposed method and our previous method. The results show that this proposed method achieves better sequencing accuracy with higher ranked correct sequences and less computational time. This paper proposed an advanced de novo sequencing method for HCD and ETD spectra pair and used information from spectra libraries and significant improved previous similar methods.

Histopathological Breast Cancer Image Classification by Deep Neural Network Techniques Guided by Local Clustering.

PubMed

Nahid, Abdullah-Al; Mehrabi, Mohamad Ali; Kong, Yinan

2018-01-01

Breast Cancer is a serious threat and one of the largest causes of death of women throughout the world. The identification of cancer largely depends on digital biomedical photography analysis such as histopathological images by doctors and physicians. Analyzing histopathological images is a nontrivial task, and decisions from investigation of these kinds of images always require specialised knowledge. However, Computer Aided Diagnosis (CAD) techniques can help the doctor make more reliable decisions. The state-of-the-art Deep Neural Network (DNN) has been recently introduced for biomedical image analysis. Normally each image contains structural and statistical information. This paper classifies a set of biomedical breast cancer images (BreakHis dataset) using novel DNN techniques guided by structural and statistical information derived from the images. Specifically a Convolutional Neural Network (CNN), a Long-Short-Term-Memory (LSTM), and a combination of CNN and LSTM are proposed for breast cancer image classification. Softmax and Support Vector Machine (SVM) layers have been used for the decision-making stage after extracting features utilising the proposed novel DNN models. In this experiment the best Accuracy value of 91.00% is achieved on the 200x dataset, the best Precision value 96.00% is achieved on the 40x dataset, and the best F -Measure value is achieved on both the 40x and 100x datasets.

A Merged Dataset for Solar Probe Plus FIELDS Magnetometers

NASA Astrophysics Data System (ADS)

Bowen, T. A.; Dudok de Wit, T.; Bale, S. D.; Revillet, C.; MacDowall, R. J.; Sheppard, D.

2016-12-01

The Solar Probe Plus FIELDS experiment will observe turbulent magnetic fluctuations deep in the inner heliosphere. The FIELDS magnetometer suite implements a set of three magnetometers: two vector DC fluxgate magnetometers (MAGs), sensitive from DC- 100Hz, as well as a vector search coil magnetometer (SCM), sensitive from 10Hz-50kHz. Single axis measurements are additionally made up to 1MHz. To study the full range of observations, we propose merging data from the individual magnetometers into a single dataset. A merged dataset will improve the quality of observations in the range of frequencies observed by both magnetometers ( 10-100 Hz). Here we present updates on the individual MAG and SCM calibrations as well as our results on generating a cross-calibrated and merged dataset.

SPHINX--an algorithm for taxonomic binning of metagenomic sequences.

PubMed

Mohammed, Monzoorul Haque; Ghosh, Tarini Shankar; Singh, Nitin Kumar; Mande, Sharmila S

2011-01-01

Compared with composition-based binning algorithms, the binning accuracy and specificity of alignment-based binning algorithms is significantly higher. However, being alignment-based, the latter class of algorithms require enormous amount of time and computing resources for binning huge metagenomic datasets. The motivation was to develop a binning approach that can analyze metagenomic datasets as rapidly as composition-based approaches, but nevertheless has the accuracy and specificity of alignment-based algorithms. This article describes a hybrid binning approach (SPHINX) that achieves high binning efficiency by utilizing the principles of both 'composition'- and 'alignment'-based binning algorithms. Validation results with simulated sequence datasets indicate that SPHINX is able to analyze metagenomic sequences as rapidly as composition-based algorithms. Furthermore, the binning efficiency (in terms of accuracy and specificity of assignments) of SPHINX is observed to be comparable with results obtained using alignment-based algorithms. A web server for the SPHINX algorithm is available at http://metagenomics.atc.tcs.com/SPHINX/.

Promoter classifier: software package for promoter database analysis.

PubMed

Gershenzon, Naum I; Ioshikhes, Ilya P

2005-01-01

Promoter Classifier is a package of seven stand-alone Windows-based C++ programs allowing the following basic manipulations with a set of promoter sequences: (i) calculation of positional distributions of nucleotides averaged over all promoters of the dataset; (ii) calculation of the averaged occurrence frequencies of the transcription factor binding sites and their combinations; (iii) division of the dataset into subsets of sequences containing or lacking certain promoter elements or combinations; (iv) extraction of the promoter subsets containing or lacking CpG islands around the transcription start site; and (v) calculation of spatial distributions of the promoter DNA stacking energy and bending stiffness. All programs have a user-friendly interface and provide the results in a convenient graphical form. The Promoter Classifier package is an effective tool for various basic manipulations with eukaryotic promoter sequences that usually are necessary for analysis of large promoter datasets. The program Promoter Divider is described in more detail as a representative component of the package.

Identification of optimum sequencing depth especially for de novo genome assembly of small genomes using next generation sequencing data.

PubMed

Desai, Aarti; Marwah, Veer Singh; Yadav, Akshay; Jha, Vineet; Dhaygude, Kishor; Bangar, Ujwala; Kulkarni, Vivek; Jere, Abhay

2013-01-01

Next Generation Sequencing (NGS) is a disruptive technology that has found widespread acceptance in the life sciences research community. The high throughput and low cost of sequencing has encouraged researchers to undertake ambitious genomic projects, especially in de novo genome sequencing. Currently, NGS systems generate sequence data as short reads and de novo genome assembly using these short reads is computationally very intensive. Due to lower cost of sequencing and higher throughput, NGS systems now provide the ability to sequence genomes at high depth. However, currently no report is available highlighting the impact of high sequence depth on genome assembly using real data sets and multiple assembly algorithms. Recently, some studies have evaluated the impact of sequence coverage, error rate and average read length on genome assembly using multiple assembly algorithms, however, these evaluations were performed using simulated datasets. One limitation of using simulated datasets is that variables such as error rates, read length and coverage which are known to impact genome assembly are carefully controlled. Hence, this study was undertaken to identify the minimum depth of sequencing required for de novo assembly for different sized genomes using graph based assembly algorithms and real datasets. Illumina reads for E.coli (4.6 MB) S.kudriavzevii (11.18 MB) and C.elegans (100 MB) were assembled using SOAPdenovo, Velvet, ABySS, Meraculous and IDBA-UD. Our analysis shows that 50X is the optimum read depth for assembling these genomes using all assemblers except Meraculous which requires 100X read depth. Moreover, our analysis shows that de novo assembly from 50X read data requires only 6-40 GB RAM depending on the genome size and assembly algorithm used. We believe that this information can be extremely valuable for researchers in designing experiments and multiplexing which will enable optimum utilization of sequencing as well as analysis resources.

Transposable element-associated microRNA hairpins produce 21-nt sRNAs integrated into typical microRNA pathways in rice

PubMed Central

Ou-Yang, Fangqian; Luo, Qing-Jun; Zhang, Yue; Richardson, Casey R.; Jiang, Yingwen; Rock, Christopher D.

2013-01-01

microRNAs (miRNAs) are a class of small RNAs (sRNAs) of ~21 nucleotides (nt) in length processed from foldback hairpins by DICER-LIKE1 (DCL1) or DCL4. They regulate the expression of target mRNAs by base pairing through RNA-Induced Silencing Complex (RISC). In the RISC, ARGONAUTE1 (AGO1) is the key protein that cleaves miRNA targets at position ten of a miRNA:target duplex. The authenticity of many annotated rice miRNA hairpins is under debate because of their homology to repeat sequences. Some of them, like miR1884b, have been removed from the current release of miRBase based on incomplete information. In this study, we investigated the association of transposable element (TE)-derived miRNAs with typical miRNA pathways (DCL1/4- and AGO1-dependent) using publicly available deep sequencing datasets. Seven miRNA hairpins with 13 unique sRNAs were specifically enriched in AGO1 immunoprecipitation samples and relatively reduced in DCL1/4 knockdown genotypes. Interestingly, these species are ~21-nt long, instead of 24-nt as annotated in miRBase and the literature. Their expression profiles meet current criteria for functional annotation of miRNAs. In addition, diagnostic cleavage tags were found in degradome datasets for predicted target mRNAs. Most of these miRNA hairpins share significant homology with miniature inverted-repeat transposable elements (MITEs), one type of abundant DNA transposons in rice. Finally, the root-specific production of a 24 nt miRNA-like sRNA was confirmed by RNA blot for a novel EST that maps to the 3'-UTR of a candidate pseudogene showing extensive sequence homology to miR1884b hairpin. Our data are consistent with the hypothesis that TEs can serve as a driving force for the evolution of some MIRNAs, where co-opting of DICER-LIKE1/4 processing and integration into AGO1 could exapt transcribed TE-associated hairpins into typical miRNA pathways. PMID:23420033

Genovo: De Novo Assembly for Metagenomes

NASA Astrophysics Data System (ADS)

Laserson, Jonathan; Jojic, Vladimir; Koller, Daphne

Next-generation sequencing technologies produce a large number of noisy reads from the DNA in a sample. Metagenomics and population sequencing aim to recover the genomic sequences of the species in the sample, which could be of high diversity. Methods geared towards single sequence reconstruction are not sensitive enough when applied in this setting. We introduce a generative probabilistic model of read generation from environmental samples and present Genovo, a novel de novo sequence assembler that discovers likely sequence reconstructions under the model. A Chinese restaurant process prior accounts for the unknown number of genomes in the sample. Inference is made by applying a series of hill-climbing steps iteratively until convergence. We compare the performance of Genovo to three other short read assembly programs across one synthetic dataset and eight metagenomic datasets created using the 454 platform, the largest of which has 311k reads. Genovo's reconstructions cover more bases and recover more genes than the other methods, and yield a higher assembly score.

Sequence Data for Clostridium autoethanogenum using Three Generations of Sequencing Technologies

DOE PAGES

Utturkar, Sagar M.; Klingeman, Dawn Marie; Bruno-Barcena, José M.; ...

2015-04-14

During the past decade, DNA sequencing output has been mostly dominated by the second generation sequencing platforms which are characterized by low cost, high throughput and shorter read lengths for example, Illumina. The emergence and development of so called third generation sequencing platforms such as PacBio has permitted exceptionally long reads (over 20 kb) to be generated. Due to read length increases, algorithm improvements and hybrid assembly approaches, the concept of one chromosome, one contig and automated finishing of microbial genomes is now a realistic and achievable task for many microbial laboratories. In this paper, we describe high quality sequencemore » datasets which span three generations of sequencing technologies, containing six types of data from four NGS platforms and originating from a single microorganism, Clostridium autoethanogenum. The dataset reported here will be useful for the scientific community to evaluate upcoming NGS platforms, enabling comparison of existing and novel bioinformatics approaches and will encourage interest in the development of innovative experimental and computational methods for NGS data.« less

RaptorX-Property: a web server for protein structure property prediction.

PubMed

Wang, Sheng; Li, Wei; Liu, Shiwang; Xu, Jinbo

2016-07-08

RaptorX Property (http://raptorx2.uchicago.edu/StructurePropertyPred/predict/) is a web server predicting structure property of a protein sequence without using any templates. It outperforms other servers, especially for proteins without close homologs in PDB or with very sparse sequence profile (i.e. carries little evolutionary information). This server employs a powerful in-house deep learning model DeepCNF (Deep Convolutional Neural Fields) to predict secondary structure (SS), solvent accessibility (ACC) and disorder regions (DISO). DeepCNF not only models complex sequence-structure relationship by a deep hierarchical architecture, but also interdependency between adjacent property labels. Our experimental results show that, tested on CASP10, CASP11 and the other benchmarks, this server can obtain ∼84% Q3 accuracy for 3-state SS, ∼72% Q8 accuracy for 8-state SS, ∼66% Q3 accuracy for 3-state solvent accessibility, and ∼0.89 area under the ROC curve (AUC) for disorder prediction. © The Author(s) 2016. Published by Oxford University Press on behalf of Nucleic Acids Research.

Deep sequencing analysis of viral infection and evolution allows rapid and detailed characterization of viral mutant spectrum.

PubMed

Isakov, Ofer; Bordería, Antonio V; Golan, David; Hamenahem, Amir; Celniker, Gershon; Yoffe, Liron; Blanc, Hervé; Vignuzzi, Marco; Shomron, Noam

2015-07-01

The study of RNA virus populations is a challenging task. Each population of RNA virus is composed of a collection of different, yet related genomes often referred to as mutant spectra or quasispecies. Virologists using deep sequencing technologies face major obstacles when studying virus population dynamics, both experimentally and in natural settings due to the relatively high error rates of these technologies and the lack of high performance pipelines. In order to overcome these hurdles we developed a computational pipeline, termed ViVan (Viral Variance Analysis). ViVan is a complete pipeline facilitating the identification, characterization and comparison of sequence variance in deep sequenced virus populations. Applying ViVan on deep sequenced data obtained from samples that were previously characterized by more classical approaches, we uncovered novel and potentially crucial aspects of virus populations. With our experimental work, we illustrate how ViVan can be used for studies ranging from the more practical, detection of resistant mutations and effects of antiviral treatments, to the more theoretical temporal characterization of the population in evolutionary studies. Freely available on the web at http://www.vivanbioinfo.org : nshomron@post.tau.ac.il Supplementary data are available at Bioinformatics online. © The Author 2015. Published by Oxford University Press.

Prediction of Drug-Target Interaction Networks from the Integration of Protein Sequences and Drug Chemical Structures.

PubMed

Meng, Fan-Rong; You, Zhu-Hong; Chen, Xing; Zhou, Yong; An, Ji-Yong

2017-07-05

Knowledge of drug-target interaction (DTI) plays an important role in discovering new drug candidates. Unfortunately, there are unavoidable shortcomings; including the time-consuming and expensive nature of the experimental method to predict DTI. Therefore, it motivates us to develop an effective computational method to predict DTI based on protein sequence. In the paper, we proposed a novel computational approach based on protein sequence, namely PDTPS (Predicting Drug Targets with Protein Sequence) to predict DTI. The PDTPS method combines Bi-gram probabilities (BIGP), Position Specific Scoring Matrix (PSSM), and Principal Component Analysis (PCA) with Relevance Vector Machine (RVM). In order to evaluate the prediction capacity of the PDTPS, the experiment was carried out on enzyme, ion channel, GPCR, and nuclear receptor datasets by using five-fold cross-validation tests. The proposed PDTPS method achieved average accuracy of 97.73%, 93.12%, 86.78%, and 87.78% on enzyme, ion channel, GPCR and nuclear receptor datasets, respectively. The experimental results showed that our method has good prediction performance. Furthermore, in order to further evaluate the prediction performance of the proposed PDTPS method, we compared it with the state-of-the-art support vector machine (SVM) classifier on enzyme and ion channel datasets, and other exiting methods on four datasets. The promising comparison results further demonstrate that the efficiency and robust of the proposed PDTPS method. This makes it a useful tool and suitable for predicting DTI, as well as other bioinformatics tasks.

SVM-PB-Pred: SVM based protein block prediction method using sequence profiles and secondary structures.

PubMed

Suresh, V; Parthasarathy, S

2014-01-01

We developed a support vector machine based web server called SVM-PB-Pred, to predict the Protein Block for any given amino acid sequence. The input features of SVM-PB-Pred include i) sequence profiles (PSSM) and ii) actual secondary structures (SS) from DSSP method or predicted secondary structures from NPS@ and GOR4 methods. There were three combined input features PSSM+SS(DSSP), PSSM+SS(NPS@) and PSSM+SS(GOR4) used to test and train the SVM models. Similarly, four datasets RS90, DB433, LI1264 and SP1577 were used to develop the SVM models. These four SVM models developed were tested using three different benchmarking tests namely; (i) self consistency, (ii) seven fold cross validation test and (iii) independent case test. The maximum possible prediction accuracy of ~70% was observed in self consistency test for the SVM models of both LI1264 and SP1577 datasets, where PSSM+SS(DSSP) input features was used to test. The prediction accuracies were reduced to ~53% for PSSM+SS(NPS@) and ~43% for PSSM+SS(GOR4) in independent case test, for the SVM models of above two same datasets. Using our method, it is possible to predict the protein block letters for any query protein sequence with ~53% accuracy, when the SP1577 dataset and predicted secondary structure from NPS@ server were used. The SVM-PB-Pred server can be freely accessed through http://bioinfo.bdu.ac.in/~svmpbpred.

Predicting protein subcellular locations using hierarchical ensemble of Bayesian classifiers based on Markov chains.

PubMed

Bulashevska, Alla; Eils, Roland

2006-06-14

The subcellular location of a protein is closely related to its function. It would be worthwhile to develop a method to predict the subcellular location for a given protein when only the amino acid sequence of the protein is known. Although many efforts have been made to predict subcellular location from sequence information only, there is the need for further research to improve the accuracy of prediction. A novel method called HensBC is introduced to predict protein subcellular location. HensBC is a recursive algorithm which constructs a hierarchical ensemble of classifiers. The classifiers used are Bayesian classifiers based on Markov chain models. We tested our method on six various datasets; among them are Gram-negative bacteria dataset, data for discriminating outer membrane proteins and apoptosis proteins dataset. We observed that our method can predict the subcellular location with high accuracy. Another advantage of the proposed method is that it can improve the accuracy of the prediction of some classes with few sequences in training and is therefore useful for datasets with imbalanced distribution of classes. This study introduces an algorithm which uses only the primary sequence of a protein to predict its subcellular location. The proposed recursive scheme represents an interesting methodology for learning and combining classifiers. The method is computationally efficient and competitive with the previously reported approaches in terms of prediction accuracies as empirical results indicate. The code for the software is available upon request.

LS³: A Method for Improving Phylogenomic Inferences When Evolutionary Rates Are Heterogeneous among Taxa.

PubMed

Rivera-Rivera, Carlos J; Montoya-Burgos, Juan I

2016-06-01

Phylogenetic inference artifacts can occur when sequence evolution deviates from assumptions made by the models used to analyze them. The combination of strong model assumption violations and highly heterogeneous lineage evolutionary rates can become problematic in phylogenetic inference, and lead to the well-described long-branch attraction (LBA) artifact. Here, we define an objective criterion for assessing lineage evolutionary rate heterogeneity among predefined lineages: the result of a likelihood ratio test between a model in which the lineages evolve at the same rate (homogeneous model) and a model in which different lineage rates are allowed (heterogeneous model). We implement this criterion in the algorithm Locus Specific Sequence Subsampling (LS³), aimed at reducing the effects of LBA in multi-gene datasets. For each gene, LS³ sequentially removes the fastest-evolving taxon of the ingroup and tests for lineage rate homogeneity until all lineages have uniform evolutionary rates. The sequences excluded from the homogeneously evolving taxon subset are flagged as potentially problematic. The software implementation provides the user with the possibility to remove the flagged sequences for generating a new concatenated alignment. We tested LS³ with simulations and two real datasets containing LBA artifacts: a nucleotide dataset regarding the position of Glires within mammals and an amino-acid dataset concerning the position of nematodes within bilaterians. The initially incorrect phylogenies were corrected in all cases upon removing data flagged by LS³. © The Author 2016. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution.

Fine Dissection of Human Mitochondrial DNA Haplogroup HV Lineages Reveals Paleolithic Signatures from European Glacial Refugia

PubMed Central

Sarno, Stefania; Sevini, Federica; Vianello, Dario; Tamm, Erika; Metspalu, Ene; van Oven, Mannis; Hübner, Alexander; Sazzini, Marco; Franceschi, Claudio; Pettener, Davide; Luiselli, Donata

2015-01-01

Genetic signatures from the Paleolithic inhabitants of Eurasia can be traced from the early divergent mitochondrial DNA lineages still present in contemporary human populations. Previous studies already suggested a pre-Neolithic diffusion of mitochondrial haplogroup HV*(xH,V) lineages, a relatively rare class of mtDNA types that includes parallel branches mainly distributed across Europe and West Asia with a certain degree of structure. Up till now, variation within haplogroup HV was addressed mainly by analyzing sequence data from the mtDNA control region, except for specific sub-branches, such as HV4 or the widely distributed haplogroups H and V. In this study, we present a revised HV topology based on full mtDNA genome data, and we include a comprehensive dataset consisting of 316 complete mtDNA sequences including 60 new samples from the Italian peninsula, a previously underrepresented geographic area. We highlight points of instability in the particular topology of this haplogroup, reconstructed with BEAST-generated trees and networks. We also confirm a major lineage expansion that probably followed the Late Glacial Maximum and preceded Neolithic population movements. We finally observe that Italy harbors a reservoir of mtDNA diversity, with deep-rooting HV lineages often related to sequences present in the Caucasus and the Middle East. The resulting hypothesis of a glacial refugium in Southern Italy has implications for the understanding of late Paleolithic population movements and is discussed within the archaeological cultural shifts occurred over the entire continent. PMID:26640946

A Mitogenomic Phylogeny of Living Primates

PubMed Central

Finstermeier, Knut; Zinner, Dietmar; Brameier, Markus; Meyer, Matthias; Kreuz, Eva; Hofreiter, Michael; Roos, Christian

2013-01-01

Primates, the mammalian order including our own species, comprise 480 species in 78 genera. Thus, they represent the third largest of the 18 orders of eutherian mammals. Although recent phylogenetic studies on primates are increasingly built on molecular datasets, most of these studies have focused on taxonomic subgroups within the order. Complete mitochondrial (mt) genomes have proven to be extremely useful in deciphering within-order relationships even up to deep nodes. Using 454 sequencing, we sequenced 32 new complete mt genomes adding 20 previously not represented genera to the phylogenetic reconstruction of the primate tree. With 13 new sequences, the number of complete mt genomes within the parvorder Platyrrhini was widely extended, resulting in a largely resolved branching pattern among New World monkey families. We added 10 new Strepsirrhini mt genomes to the 15 previously available ones, thus almost doubling the number of mt genomes within this clade. Our data allow precise date estimates of all nodes and offer new insights into primate evolution. One major result is a relatively young date for the most recent common ancestor of all living primates which was estimated to 66-69 million years ago, suggesting that the divergence of extant primates started close to the K/T-boundary. Although some relationships remain unclear, the large number of mt genomes used allowed us to reconstruct a robust primate phylogeny which is largely in agreement with previous publications. Finally, we show that mt genomes are a useful tool for resolving primate phylogenetic relationships on various taxonomic levels. PMID:23874967

The Tara Oceans voyage reveals global diversity and distribution patterns of marine planktonic ciliates

PubMed Central

Gimmler, Anna; Korn, Ralf; de Vargas, Colomban; Audic, Stéphane; Stoeck, Thorsten

2016-01-01

Illumina reads of the SSU-rDNA-V9 region obtained from the circumglobal Tara Oceans expedition allow the investigation of protistan plankton diversity patterns on a global scale. We analyzed 6,137,350 V9-amplicons from ocean surface waters and the deep chlorophyll maximum, which were taxonomically assigned to the phylum Ciliophora. For open ocean samples global planktonic ciliate diversity is relatively low (ca. 1,300 observed and predicted ciliate OTUs). We found that 17% of all detected ciliate OTUs occurred in all oceanic regions under study. On average, local ciliate OTU richness represented 27% of the global ciliate OTU richness, indicating that a large proportion of ciliates is widely distributed. Yet, more than half of these OTUs shared <90% sequence similarity with reference sequences of described ciliates. While alpha-diversity measures (richness and exp(Shannon H)) are hardly affected by contemporary environmental conditions, species (OTU) turnover and community similarity (β-diversity) across taxonomic groups showed strong correlation to environmental parameters. Logistic regression models predicted significant correlations between the occurrence of specific ciliate genera and individual nutrients, the oceanic carbonate system and temperature. Planktonic ciliates displayed distinct vertical distributions relative to chlorophyll a. In contrast, the Tara Oceans dataset did not reveal any evidence that latitude is structuring ciliate communities. PMID:27633177

MINTmap: fast and exhaustive profiling of nuclear and mitochondrial tRNA fragments from short RNA-seq data

PubMed Central

Loher, Phillipe; Telonis, Aristeidis G.; Rigoutsos, Isidore

2017-01-01

Transfer RNA fragments (tRFs) are an established class of constitutive regulatory molecules that arise from precursor and mature tRNAs. RNA deep sequencing (RNA-seq) has greatly facilitated the study of tRFs. However, the repeat nature of the tRNA templates and the idiosyncrasies of tRNA sequences necessitate the development and use of methodologies that differ markedly from those used to analyze RNA-seq data when studying microRNAs (miRNAs) or messenger RNAs (mRNAs). Here we present MINTmap (for MItochondrial and Nuclear TRF mapping), a method and a software package that was developed specifically for the quick, deterministic and exhaustive identification of tRFs in short RNA-seq datasets. In addition to identifying them, MINTmap is able to unambiguously calculate and report both raw and normalized abundances for the discovered tRFs. Furthermore, to ensure specificity, MINTmap identifies the subset of discovered tRFs that could be originating outside of tRNA space and flags them as candidate false positives. Our comparative analysis shows that MINTmap exhibits superior sensitivity and specificity to other available methods while also being exceptionally fast. The MINTmap codes are available through https://github.com/TJU-CMC-Org/MINTmap/ under an open source GNU GPL v3.0 license. PMID:28220888

Identification of ribonucleotide reductase mutation causing temperature-sensitivity of herpes simplex virus isolates from whitlow by deep sequencing.

PubMed

Daikoku, Tohru; Oyama, Yukari; Yajima, Misako; Sekizuka, Tsuyoshi; Kuroda, Makoto; Shimada, Yuka; Takehara, Kazuhiko; Miwa, Naoko; Okuda, Tomoko; Sata, Tetsutaro; Shiraki, Kimiyasu

2015-06-01

Herpes simplex virus 2 caused a genital ulcer, and a secondary herpetic whitlow appeared during acyclovir therapy. The secondary and recurrent whitlow isolates were acyclovir-resistant and temperature-sensitive in contrast to a genital isolate. We identified the ribonucleotide reductase mutation responsible for temperature-sensitivity by deep-sequencing analysis.

Detecting exact breakpoints of deletions with diversity in hepatitis B viral genomic DNA from next-generation sequencing data.

PubMed

Cheng, Ji-Hong; Liu, Wen-Chun; Chang, Ting-Tsung; Hsieh, Sun-Yuan; Tseng, Vincent S

2017-10-01

Many studies have suggested that deletions of Hepatitis B Viral (HBV) are associated with the development of progressive liver diseases, even ultimately resulting in hepatocellular carcinoma (HCC). Among the methods for detecting deletions from next-generation sequencing (NGS) data, few methods considered the characteristics of virus, such as high evolution rates and high divergence among the different HBV genomes. Sequencing high divergence HBV genome sequences using the NGS technology outputs millions of reads. Thus, detecting exact breakpoints of deletions from these big and complex data incurs very high computational cost. We proposed a novel analytical method named VirDelect (Virus Deletion Detect), which uses split read alignment base to detect exact breakpoint and diversity variable to consider high divergence in single-end reads data, such that the computational cost can be reduced without losing accuracy. We use four simulated reads datasets and two real pair-end reads datasets of HBV genome sequence to verify VirDelect accuracy by score functions. The experimental results show that VirDelect outperforms the state-of-the-art method Pindel in terms of accuracy score for all simulated datasets and VirDelect had only two base errors even in real datasets. VirDelect is also shown to deliver high accuracy in analyzing the single-end read data as well as pair-end data. VirDelect can serve as an effective and efficient bioinformatics tool for physiologists with high accuracy and efficient performance and applicable to further analysis with characteristics similar to HBV on genome length and high divergence. The software program of VirDelect can be downloaded at https://sourceforge.net/projects/virdelect/. Copyright © 2017. Published by Elsevier Inc.

miRCat2: accurate prediction of plant and animal microRNAs from next-generation sequencing datasets

PubMed Central

Paicu, Claudia; Mohorianu, Irina; Stocks, Matthew; Xu, Ping; Coince, Aurore; Billmeier, Martina; Dalmay, Tamas; Moulton, Vincent; Moxon, Simon

2017-01-01

Abstract Motivation MicroRNAs are a class of ∼21–22 nt small RNAs which are excised from a stable hairpin-like secondary structure. They have important gene regulatory functions and are involved in many pathways including developmental timing, organogenesis and development in eukaryotes. There are several computational tools for miRNA detection from next-generation sequencing datasets. However, many of these tools suffer from high false positive and false negative rates. Here we present a novel miRNA prediction algorithm, miRCat2. miRCat2 incorporates a new entropy-based approach to detect miRNA loci, which is designed to cope with the high sequencing depth of current next-generation sequencing datasets. It has a user-friendly interface and produces graphical representations of the hairpin structure and plots depicting the alignment of sequences on the secondary structure. Results We test miRCat2 on a number of animal and plant datasets and present a comparative analysis with miRCat, miRDeep2, miRPlant and miReap. We also use mutants in the miRNA biogenesis pathway to evaluate the predictions of these tools. Results indicate that miRCat2 has an improved accuracy compared with other methods tested. Moreover, miRCat2 predicts several new miRNAs that are differentially expressed in wild-type versus mutants in the miRNA biogenesis pathway. Availability and Implementation miRCat2 is part of the UEA small RNA Workbench and is freely available from http://srna-workbench.cmp.uea.ac.uk/. Contact v.moulton@uea.ac.uk or s.moxon@uea.ac.uk Supplementary information Supplementary data are available at Bioinformatics online. PMID:28407097

Chiron: translating nanopore raw signal directly into nucleotide sequence using deep learning.

PubMed

Teng, Haotian; Cao, Minh Duc; Hall, Michael B; Duarte, Tania; Wang, Sheng; Coin, Lachlan J M

2018-05-01

Sequencing by translocating DNA fragments through an array of nanopores is a rapidly maturing technology that offers faster and cheaper sequencing than other approaches. However, accurately deciphering the DNA sequence from the noisy and complex electrical signal is challenging. Here, we report Chiron, the first deep learning model to achieve end-to-end basecalling and directly translate the raw signal to DNA sequence without the error-prone segmentation step. Trained with only a small set of 4,000 reads, we show that our model provides state-of-the-art basecalling accuracy, even on previously unseen species. Chiron achieves basecalling speeds of more than 2,000 bases per second using desktop computer graphics processing units.

Rhipicephalus microplus dataset of nonredundant raw sequence reads from 454 GS FLX sequencing of Cot-selected (Cot = 660) genomic DNA

USDA-ARS?s Scientific Manuscript database

A reassociation kinetics-based approach was used to reduce the complexity of genomic DNA from the Deutsch laboratory strain of the cattle tick, Rhipicephalus microplus, to facilitate genome sequencing. Selected genomic DNA (Cot value = 660) was sequenced using 454 GS FLX technology, resulting in 356...

Object class segmentation of RGB-D video using recurrent convolutional neural networks.

PubMed

Pavel, Mircea Serban; Schulz, Hannes; Behnke, Sven

2017-04-01

Object class segmentation is a computer vision task which requires labeling each pixel of an image with the class of the object it belongs to. Deep convolutional neural networks (DNN) are able to learn and take advantage of local spatial correlations required for this task. They are, however, restricted by their small, fixed-sized filters, which limits their ability to learn long-range dependencies. Recurrent Neural Networks (RNN), on the other hand, do not suffer from this restriction. Their iterative interpretation allows them to model long-range dependencies by propagating activity. This property is especially useful when labeling video sequences, where both spatial and temporal long-range dependencies occur. In this work, a novel RNN architecture for object class segmentation is presented. We investigate several ways to train such a network. We evaluate our models on the challenging NYU Depth v2 dataset for object class segmentation and obtain competitive results. Copyright © 2017 Elsevier Ltd. All rights reserved.

Deep transcriptome annotation enables the discovery and functional characterization of cryptic small proteins

PubMed Central

Delcourt, Vivian; Lucier, Jean-François; Gagnon, Jules; Beaudoin, Maxime C; Vanderperre, Benoît; Breton, Marc-André; Motard, Julie; Jacques, Jean-François; Brunelle, Mylène; Gagnon-Arsenault, Isabelle; Fournier, Isabelle; Ouangraoua, Aida; Hunting, Darel J; Cohen, Alan A; Landry, Christian R; Scott, Michelle S

2017-01-01

Recent functional, proteomic and ribosome profiling studies in eukaryotes have concurrently demonstrated the translation of alternative open-reading frames (altORFs) in addition to annotated protein coding sequences (CDSs). We show that a large number of small proteins could in fact be coded by these altORFs. The putative alternative proteins translated from altORFs have orthologs in many species and contain functional domains. Evolutionary analyses indicate that altORFs often show more extreme conservation patterns than their CDSs. Thousands of alternative proteins are detected in proteomic datasets by reanalysis using a database containing predicted alternative proteins. This is illustrated with specific examples, including altMiD51, a 70 amino acid mitochondrial fission-promoting protein encoded in MiD51/Mief1/SMCR7L, a gene encoding an annotated protein promoting mitochondrial fission. Our results suggest that many genes are multicoding genes and code for a large protein and one or several small proteins. PMID:29083303

Systems Biology of Metabolic Regulation by Estrogen Receptor Signaling in Breast Cancer.

PubMed

Zhao, Yiru Chen; Madak Erdogan, Zeynep

2016-03-17

With the advent of the -omics approaches our understanding of the chronic diseases like cancer and metabolic syndrome has improved. However, effective mining of the information in the large-scale datasets that are obtained from gene expression microarrays, deep sequencing experiments or metabolic profiling is essential to uncover and then effectively target the critical regulators of diseased cell phenotypes. Estrogen Receptor α (ERα) is one of the master transcription factors regulating the gene programs that are important for estrogen responsive breast cancers. In order to understand to role of ERα signaling in breast cancer metabolism we utilized transcriptomic, cistromic and metabolomic data from MCF-7 cells treated with estradiol. In this report we described generation of samples for RNA-Seq, ChIP-Seq and metabolomics experiments and the integrative computational analysis of the obtained data. This approach is useful in delineating novel molecular mechanisms and gene regulatory circuits that are regulated by a particular transcription factor which impacts metabolism of normal or diseased cells.

Fine-grained leukocyte classification with deep residual learning for microscopic images.

PubMed

Qin, Feiwei; Gao, Nannan; Peng, Yong; Wu, Zizhao; Shen, Shuying; Grudtsin, Artur

2018-08-01

Leukocyte classification and cytometry have wide applications in medical domain, previous researches usually exploit machine learning techniques to classify leukocytes automatically. However, constrained by the past development of machine learning techniques, for example, extracting distinctive features from raw microscopic images are difficult, the widely used SVM classifier only has relative few parameters to tune, these methods cannot efficiently handle fine-grained classification cases when the white blood cells have up to 40 categories. Based on deep learning theory, a systematic study is conducted on finer leukocyte classification in this paper. A deep residual neural network based leukocyte classifier is constructed at first, which can imitate the domain expert's cell recognition process, and extract salient features robustly and automatically. Then the deep neural network classifier's topology is adjusted according to the prior knowledge of white blood cell test. After that the microscopic image dataset with almost one hundred thousand labeled leukocytes belonging to 40 categories is built, and combined training strategies are adopted to make the designed classifier has good generalization ability. The proposed deep residual neural network based classifier was tested on microscopic image dataset with 40 leukocyte categories. It achieves top-1 accuracy of 77.80%, top-5 accuracy of 98.75% during the training procedure. The average accuracy on the test set is nearly 76.84%. This paper presents a fine-grained leukocyte classification method for microscopic images, based on deep residual learning theory and medical domain knowledge. Experimental results validate the feasibility and effectiveness of our approach. Extended experiments support that the fine-grained leukocyte classifier could be used in real medical applications, assist doctors in diagnosing diseases, reduce human power significantly. Copyright © 2018 Elsevier B.V. All rights reserved.

Automatically identifying, counting, and describing wild animals in camera-trap images with deep learning.

PubMed

Norouzzadeh, Mohammad Sadegh; Nguyen, Anh; Kosmala, Margaret; Swanson, Alexandra; Palmer, Meredith S; Packer, Craig; Clune, Jeff

2018-06-19

Having accurate, detailed, and up-to-date information about the location and behavior of animals in the wild would improve our ability to study and conserve ecosystems. We investigate the ability to automatically, accurately, and inexpensively collect such data, which could help catalyze the transformation of many fields of ecology, wildlife biology, zoology, conservation biology, and animal behavior into "big data" sciences. Motion-sensor "camera traps" enable collecting wildlife pictures inexpensively, unobtrusively, and frequently. However, extracting information from these pictures remains an expensive, time-consuming, manual task. We demonstrate that such information can be automatically extracted by deep learning, a cutting-edge type of artificial intelligence. We train deep convolutional neural networks to identify, count, and describe the behaviors of 48 species in the 3.2 million-image Snapshot Serengeti dataset. Our deep neural networks automatically identify animals with >93.8% accuracy, and we expect that number to improve rapidly in years to come. More importantly, if our system classifies only images it is confident about, our system can automate animal identification for 99.3% of the data while still performing at the same 96.6% accuracy as that of crowdsourced teams of human volunteers, saving >8.4 y (i.e., >17,000 h at 40 h/wk) of human labeling effort on this 3.2 million-image dataset. Those efficiency gains highlight the importance of using deep neural networks to automate data extraction from camera-trap images, reducing a roadblock for this widely used technology. Our results suggest that deep learning could enable the inexpensive, unobtrusive, high-volume, and even real-time collection of a wealth of information about vast numbers of animals in the wild. Copyright © 2018 the Author(s). Published by PNAS.

Evaluation of Nine Somatic Variant Callers for Detection of Somatic Mutations in Exome and Targeted Deep Sequencing Data.

PubMed

Krøigård, Anne Bruun; Thomassen, Mads; Lænkholm, Anne-Vibeke; Kruse, Torben A; Larsen, Martin Jakob

2016-01-01

Next generation sequencing is extensively applied to catalogue somatic mutations in cancer, in research settings and increasingly in clinical settings for molecular diagnostics, guiding therapy decisions. Somatic variant callers perform paired comparisons of sequencing data from cancer tissue and matched normal tissue in order to detect somatic mutations. The advent of many new somatic variant callers creates a need for comparison and validation of the tools, as no de facto standard for detection of somatic mutations exists and only limited comparisons have been reported. We have performed a comprehensive evaluation using exome sequencing and targeted deep sequencing data of paired tumor-normal samples from five breast cancer patients to evaluate the performance of nine publicly available somatic variant callers: EBCall, Mutect, Seurat, Shimmer, Indelocator, Somatic Sniper, Strelka, VarScan 2 and Virmid for the detection of single nucleotide mutations and small deletions and insertions. We report a large variation in the number of calls from the nine somatic variant callers on the same sequencing data and highly variable agreement. Sequencing depth had markedly diverse impact on individual callers, as for some callers, increased sequencing depth highly improved sensitivity. For SNV calling, we report EBCall, Mutect, Virmid and Strelka to be the most reliable somatic variant callers for both exome sequencing and targeted deep sequencing. For indel calling, EBCall is superior due to high sensitivity and robustness to changes in sequencing depths.

Evaluation of Nine Somatic Variant Callers for Detection of Somatic Mutations in Exome and Targeted Deep Sequencing Data

PubMed Central

Krøigård, Anne Bruun; Thomassen, Mads; Lænkholm, Anne-Vibeke; Kruse, Torben A.; Larsen, Martin Jakob

2016-01-01

Next generation sequencing is extensively applied to catalogue somatic mutations in cancer, in research settings and increasingly in clinical settings for molecular diagnostics, guiding therapy decisions. Somatic variant callers perform paired comparisons of sequencing data from cancer tissue and matched normal tissue in order to detect somatic mutations. The advent of many new somatic variant callers creates a need for comparison and validation of the tools, as no de facto standard for detection of somatic mutations exists and only limited comparisons have been reported. We have performed a comprehensive evaluation using exome sequencing and targeted deep sequencing data of paired tumor-normal samples from five breast cancer patients to evaluate the performance of nine publicly available somatic variant callers: EBCall, Mutect, Seurat, Shimmer, Indelocator, Somatic Sniper, Strelka, VarScan 2 and Virmid for the detection of single nucleotide mutations and small deletions and insertions. We report a large variation in the number of calls from the nine somatic variant callers on the same sequencing data and highly variable agreement. Sequencing depth had markedly diverse impact on individual callers, as for some callers, increased sequencing depth highly improved sensitivity. For SNV calling, we report EBCall, Mutect, Virmid and Strelka to be the most reliable somatic variant callers for both exome sequencing and targeted deep sequencing. For indel calling, EBCall is superior due to high sensitivity and robustness to changes in sequencing depths. PMID:27002637

S-CNN: Subcategory-aware convolutional networks for object detection.

PubMed

Chen, Tao; Lu, Shijian; Fan, Jiayuan

2017-09-26

The marriage between the deep convolutional neural network (CNN) and region proposals has made breakthroughs for object detection in recent years. While the discriminative object features are learned via a deep CNN for classification, the large intra-class variation and deformation still limit the performance of the CNN based object detection. We propose a subcategory-aware CNN (S-CNN) to solve the object intra-class variation problem. In the proposed technique, the training samples are first grouped into multiple subcategories automatically through a novel instance sharing maximum margin clustering process. A multi-component Aggregated Channel Feature (ACF) detector is then trained to produce more latent training samples, where each ACF component corresponds to one clustered subcategory. The produced latent samples together with their subcategory labels are further fed into a CNN classifier to filter out false proposals for object detection. An iterative learning algorithm is designed for the joint optimization of image subcategorization, multi-component ACF detector, and subcategory-aware CNN classifier. Experiments on INRIA Person dataset, Pascal VOC 2007 dataset and MS COCO dataset show that the proposed technique clearly outperforms the state-of-the-art methods for generic object detection.

Design of multiple sequence alignment algorithms on parallel, distributed memory supercomputers.

PubMed

Church, Philip C; Goscinski, Andrzej; Holt, Kathryn; Inouye, Michael; Ghoting, Amol; Makarychev, Konstantin; Reumann, Matthias

2011-01-01

The challenge of comparing two or more genomes that have undergone recombination and substantial amounts of segmental loss and gain has recently been addressed for small numbers of genomes. However, datasets of hundreds of genomes are now common and their sizes will only increase in the future. Multiple sequence alignment of hundreds of genomes remains an intractable problem due to quadratic increases in compute time and memory footprint. To date, most alignment algorithms are designed for commodity clusters without parallelism. Hence, we propose the design of a multiple sequence alignment algorithm on massively parallel, distributed memory supercomputers to enable research into comparative genomics on large data sets. Following the methodology of the sequential progressiveMauve algorithm, we design data structures including sequences and sorted k-mer lists on the IBM Blue Gene/P supercomputer (BG/P). Preliminary results show that we can reduce the memory footprint so that we can potentially align over 250 bacterial genomes on a single BG/P compute node. We verify our results on a dataset of E.coli, Shigella and S.pneumoniae genomes. Our implementation returns results matching those of the original algorithm but in 1/2 the time and with 1/4 the memory footprint for scaffold building. In this study, we have laid the basis for multiple sequence alignment of large-scale datasets on a massively parallel, distributed memory supercomputer, thus enabling comparison of hundreds instead of a few genome sequences within reasonable time.

Deep learning in pharmacogenomics: from gene regulation to patient stratification.

PubMed

Kalinin, Alexandr A; Higgins, Gerald A; Reamaroon, Narathip; Soroushmehr, Sayedmohammadreza; Allyn-Feuer, Ari; Dinov, Ivo D; Najarian, Kayvan; Athey, Brian D

2018-05-01

This Perspective provides examples of current and future applications of deep learning in pharmacogenomics, including: identification of novel regulatory variants located in noncoding domains of the genome and their function as applied to pharmacoepigenomics; patient stratification from medical records; and the mechanistic prediction of drug response, targets and their interactions. Deep learning encapsulates a family of machine learning algorithms that has transformed many important subfields of artificial intelligence over the last decade, and has demonstrated breakthrough performance improvements on a wide range of tasks in biomedicine. We anticipate that in the future, deep learning will be widely used to predict personalized drug response and optimize medication selection and dosing, using knowledge extracted from large and complex molecular, epidemiological, clinical and demographic datasets.

Deep learning approach to bacterial colony classification.

PubMed

Zieliński, Bartosz; Plichta, Anna; Misztal, Krzysztof; Spurek, Przemysław; Brzychczy-Włoch, Monika; Ochońska, Dorota

2017-01-01

In microbiology it is diagnostically useful to recognize various genera and species of bacteria. It can be achieved using computer-aided methods, which make the recognition processes more automatic and thus significantly reduce the time necessary for the classification. Moreover, in case of diagnostic uncertainty (the misleading similarity in shape or structure of bacterial cells), such methods can minimize the risk of incorrect recognition. In this article, we apply the state of the art method for texture analysis to classify genera and species of bacteria. This method uses deep Convolutional Neural Networks to obtain image descriptors, which are then encoded and classified with Support Vector Machine or Random Forest. To evaluate this approach and to make it comparable with other approaches, we provide a new dataset of images. DIBaS dataset (Digital Image of Bacterial Species) contains 660 images with 33 different genera and species of bacteria.

Development of a Deep Learning Algorithm for Automatic Diagnosis of Diabetic Retinopathy.

PubMed

Raju, Manoj; Pagidimarri, Venkatesh; Barreto, Ryan; Kadam, Amrit; Kasivajjala, Vamsichandra; Aswath, Arun

2017-01-01

This paper mainly focuses on the deep learning application in classifying the stage of diabetic retinopathy and detecting the laterality of the eye using funduscopic images. Diabetic retinopathy is a chronic, progressive, sight-threatening disease of the retinal blood vessels. Ophthalmologists diagnose diabetic retinopathy through early funduscopic screening. Normally, there is a time delay in reporting and intervention, apart from the financial cost and risk of blindness associated with it. Using a convolutional neural network based approach for automatic diagnosis of diabetic retinopathy, we trained the prediction network on the publicly available Kaggle dataset. Approximately 35,000 images were used to train the network, which observed a sensitivity of 80.28% and a specificity of 92.29% on the validation dataset of ~53,000 images. Using 8,810 images, the network was trained for detecting the laterality of the eye and observed an accuracy of 93.28% on the validation set of 8,816 images.

Feedback control in deep drawing based on experimental datasets

NASA Astrophysics Data System (ADS)

Fischer, P.; Heingärtner, J.; Aichholzer, W.; Hortig, D.; Hora, P.

2017-09-01

In large-scale production of deep drawing parts, like in automotive industry, the effects of scattering material properties as well as warming of the tools have a significant impact on the drawing result. In the scope of the work, an approach is presented to minimize the influence of these effects on part quality by optically measuring the draw-in of each part and adjusting the settings of the press to keep the strain distribution, which is represented by the draw-in, inside a certain limit. For the design of the control algorithm, a design of experiments for in-line tests is used to quantify the influence of the blank holder force as well as the force distribution on the draw-in. The results of this experimental dataset are used to model the process behavior. Based on this model, a feedback control loop is designed. Finally, the performance of the control algorithm is validated in the production line.

Enhancement of ELDA Tracker Based on CNN Features and Adaptive Model Update.

PubMed

Gao, Changxin; Shi, Huizhang; Yu, Jin-Gang; Sang, Nong

2016-04-15

Appearance representation and the observation model are the most important components in designing a robust visual tracking algorithm for video-based sensors. Additionally, the exemplar-based linear discriminant analysis (ELDA) model has shown good performance in object tracking. Based on that, we improve the ELDA tracking algorithm by deep convolutional neural network (CNN) features and adaptive model update. Deep CNN features have been successfully used in various computer vision tasks. Extracting CNN features on all of the candidate windows is time consuming. To address this problem, a two-step CNN feature extraction method is proposed by separately computing convolutional layers and fully-connected layers. Due to the strong discriminative ability of CNN features and the exemplar-based model, we update both object and background models to improve their adaptivity and to deal with the tradeoff between discriminative ability and adaptivity. An object updating method is proposed to select the "good" models (detectors), which are quite discriminative and uncorrelated to other selected models. Meanwhile, we build the background model as a Gaussian mixture model (GMM) to adapt to complex scenes, which is initialized offline and updated online. The proposed tracker is evaluated on a benchmark dataset of 50 video sequences with various challenges. It achieves the best overall performance among the compared state-of-the-art trackers, which demonstrates the effectiveness and robustness of our tracking algorithm.

Enhancement of ELDA Tracker Based on CNN Features and Adaptive Model Update

PubMed Central

Gao, Changxin; Shi, Huizhang; Yu, Jin-Gang; Sang, Nong

2016-01-01

Appearance representation and the observation model are the most important components in designing a robust visual tracking algorithm for video-based sensors. Additionally, the exemplar-based linear discriminant analysis (ELDA) model has shown good performance in object tracking. Based on that, we improve the ELDA tracking algorithm by deep convolutional neural network (CNN) features and adaptive model update. Deep CNN features have been successfully used in various computer vision tasks. Extracting CNN features on all of the candidate windows is time consuming. To address this problem, a two-step CNN feature extraction method is proposed by separately computing convolutional layers and fully-connected layers. Due to the strong discriminative ability of CNN features and the exemplar-based model, we update both object and background models to improve their adaptivity and to deal with the tradeoff between discriminative ability and adaptivity. An object updating method is proposed to select the “good” models (detectors), which are quite discriminative and uncorrelated to other selected models. Meanwhile, we build the background model as a Gaussian mixture model (GMM) to adapt to complex scenes, which is initialized offline and updated online. The proposed tracker is evaluated on a benchmark dataset of 50 video sequences with various challenges. It achieves the best overall performance among the compared state-of-the-art trackers, which demonstrates the effectiveness and robustness of our tracking algorithm. PMID:27092505

Modeling positional effects of regulatory sequences with spline transformations increases prediction accuracy of deep neural networks

PubMed Central

Avsec, Žiga; Cheng, Jun; Gagneur, Julien

2018-01-01

Abstract Motivation Regulatory sequences are not solely defined by their nucleic acid sequence but also by their relative distances to genomic landmarks such as transcription start site, exon boundaries or polyadenylation site. Deep learning has become the approach of choice for modeling regulatory sequences because of its strength to learn complex sequence features. However, modeling relative distances to genomic landmarks in deep neural networks has not been addressed. Results Here we developed spline transformation, a neural network module based on splines to flexibly and robustly model distances. Modeling distances to various genomic landmarks with spline transformations significantly increased state-of-the-art prediction accuracy of in vivo RNA-binding protein binding sites for 120 out of 123 proteins. We also developed a deep neural network for human splice branchpoint based on spline transformations that outperformed the current best, already distance-based, machine learning model. Compared to piecewise linear transformation, as obtained by composition of rectified linear units, spline transformation yields higher prediction accuracy as well as faster and more robust training. As spline transformation can be applied to further quantities beyond distances, such as methylation or conservation, we foresee it as a versatile component in the genomics deep learning toolbox. Availability and implementation Spline transformation is implemented as a Keras layer in the CONCISE python package: https://github.com/gagneurlab/concise. Analysis code is available at https://github.com/gagneurlab/Manuscript_Avsec_Bioinformatics_2017. Contact avsec@in.tum.de or gagneur@in.tum.de Supplementary information Supplementary data are available at Bioinformatics online. PMID:29155928

Deep Convolutional Neural Networks for breast cancer screening.

PubMed

Chougrad, Hiba; Zouaki, Hamid; Alheyane, Omar

2018-04-01

Radiologists often have a hard time classifying mammography mass lesions which leads to unnecessary breast biopsies to remove suspicions and this ends up adding exorbitant expenses to an already burdened patient and health care system. In this paper we developed a Computer-aided Diagnosis (CAD) system based on deep Convolutional Neural Networks (CNN) that aims to help the radiologist classify mammography mass lesions. Deep learning usually requires large datasets to train networks of a certain depth from scratch. Transfer learning is an effective method to deal with relatively small datasets as in the case of medical images, although it can be tricky as we can easily start overfitting. In this work, we explore the importance of transfer learning and we experimentally determine the best fine-tuning strategy to adopt when training a CNN model. We were able to successfully fine-tune some of the recent, most powerful CNNs and achieved better results compared to other state-of-the-art methods which classified the same public datasets. For instance we achieved 97.35% accuracy and 0.98 AUC on the DDSM database, 95.50% accuracy and 0.97 AUC on the INbreast database and 96.67% accuracy and 0.96 AUC on the BCDR database. Furthermore, after pre-processing and normalizing all the extracted Regions of Interest (ROIs) from the full mammograms, we merged all the datasets to build one large set of images and used it to fine-tune our CNNs. The CNN model which achieved the best results, a 98.94% accuracy, was used as a baseline to build the Breast Cancer Screening Framework. To evaluate the proposed CAD system and its efficiency to classify new images, we tested it on an independent database (MIAS) and got 98.23% accuracy and 0.99 AUC. The results obtained demonstrate that the proposed framework is performant and can indeed be used to predict if the mass lesions are benign or malignant. Copyright © 2018 Elsevier B.V. All rights reserved.

A deep learning approach for the analysis of masses in mammograms with minimal user intervention.

PubMed

Dhungel, Neeraj; Carneiro, Gustavo; Bradley, Andrew P

2017-04-01

We present an integrated methodology for detecting, segmenting and classifying breast masses from mammograms with minimal user intervention. This is a long standing problem due to low signal-to-noise ratio in the visualisation of breast masses, combined with their large variability in terms of shape, size, appearance and location. We break the problem down into three stages: mass detection, mass segmentation, and mass classification. For the detection, we propose a cascade of deep learning methods to select hypotheses that are refined based on Bayesian optimisation. For the segmentation, we propose the use of deep structured output learning that is subsequently refined by a level set method. Finally, for the classification, we propose the use of a deep learning classifier, which is pre-trained with a regression to hand-crafted feature values and fine-tuned based on the annotations of the breast mass classification dataset. We test our proposed system on the publicly available INbreast dataset and compare the results with the current state-of-the-art methodologies. This evaluation shows that our system detects 90% of masses at 1 false positive per image, has a segmentation accuracy of around 0.85 (Dice index) on the correctly detected masses, and overall classifies masses as malignant or benign with sensitivity (Se) of 0.98 and specificity (Sp) of 0.7. Copyright © 2017 Elsevier B.V. All rights reserved.

The Next Era: Deep Learning in Pharmaceutical Research.

PubMed

Ekins, Sean

2016-11-01

Over the past decade we have witnessed the increasing sophistication of machine learning algorithms applied in daily use from internet searches, voice recognition, social network software to machine vision software in cameras, phones, robots and self-driving cars. Pharmaceutical research has also seen its fair share of machine learning developments. For example, applying such methods to mine the growing datasets that are created in drug discovery not only enables us to learn from the past but to predict a molecule's properties and behavior in future. The latest machine learning algorithm garnering significant attention is deep learning, which is an artificial neural network with multiple hidden layers. Publications over the last 3 years suggest that this algorithm may have advantages over previous machine learning methods and offer a slight but discernable edge in predictive performance. The time has come for a balanced review of this technique but also to apply machine learning methods such as deep learning across a wider array of endpoints relevant to pharmaceutical research for which the datasets are growing such as physicochemical property prediction, formulation prediction, absorption, distribution, metabolism, excretion and toxicity (ADME/Tox), target prediction and skin permeation, etc. We also show that there are many potential applications of deep learning beyond cheminformatics. It will be important to perform prospective testing (which has been carried out rarely to date) in order to convince skeptics that there will be benefits from investing in this technique.

Accurate segmentation of lung fields on chest radiographs using deep convolutional networks

NASA Astrophysics Data System (ADS)

Arbabshirani, Mohammad R.; Dallal, Ahmed H.; Agarwal, Chirag; Patel, Aalpan; Moore, Gregory

2017-02-01

Accurate segmentation of lung fields on chest radiographs is the primary step for computer-aided detection of various conditions such as lung cancer and tuberculosis. The size, shape and texture of lung fields are key parameters for chest X-ray (CXR) based lung disease diagnosis in which the lung field segmentation is a significant primary step. Although many methods have been proposed for this problem, lung field segmentation remains as a challenge. In recent years, deep learning has shown state of the art performance in many visual tasks such as object detection, image classification and semantic image segmentation. In this study, we propose a deep convolutional neural network (CNN) framework for segmentation of lung fields. The algorithm was developed and tested on 167 clinical posterior-anterior (PA) CXR images collected retrospectively from picture archiving and communication system (PACS) of Geisinger Health System. The proposed multi-scale network is composed of five convolutional and two fully connected layers. The framework achieved IOU (intersection over union) of 0.96 on the testing dataset as compared to manual segmentation. The suggested framework outperforms state of the art registration-based segmentation by a significant margin. To our knowledge, this is the first deep learning based study of lung field segmentation on CXR images developed on a heterogeneous clinical dataset. The results suggest that convolutional neural networks could be employed reliably for lung field segmentation.

Identifying Group-Specific Sequences for Microbial Communities Using Long k-mer Sequence Signatures

PubMed Central

Wang, Ying; Fu, Lei; Ren, Jie; Yu, Zhaoxia; Chen, Ting; Sun, Fengzhu

2018-01-01

Comparing metagenomic samples is crucial for understanding microbial communities. For different groups of microbial communities, such as human gut metagenomic samples from patients with a certain disease and healthy controls, identifying group-specific sequences offers essential information for potential biomarker discovery. A sequence that is present, or rich, in one group, but absent, or scarce, in another group is considered “group-specific” in our study. Our main purpose is to discover group-specific sequence regions between control and case groups as disease-associated markers. We developed a long k-mer (k ≥ 30 bps)-based computational pipeline to detect group-specific sequences at strain resolution free from reference sequences, sequence alignments, and metagenome-wide de novo assembly. We called our method MetaGO: Group-specific oligonucleotide analysis for metagenomic samples. An open-source pipeline on Apache Spark was developed with parallel computing. We applied MetaGO to one simulated and three real metagenomic datasets to evaluate the discriminative capability of identified group-specific markers. In the simulated dataset, 99.11% of group-specific logical 40-mers covered 98.89% disease-specific regions from the disease-associated strain. In addition, 97.90% of group-specific numerical 40-mers covered 99.61 and 96.39% of differentially abundant genome and regions between two groups, respectively. For a large-scale metagenomic liver cirrhosis (LC)-associated dataset, we identified 37,647 group-specific 40-mer features. Any one of the features can predict disease status of the training samples with the average of sensitivity and specificity higher than 0.8. The random forests classification using the top 10 group-specific features yielded a higher AUC (from ∼0.8 to ∼0.9) than that of previous studies. All group-specific 40-mers were present in LC patients, but not healthy controls. All the assembled 11 LC-specific sequences can be mapped to two strains of Veillonella parvula: UTDB1-3 and DSM2008. The experiments on the other two real datasets related to Inflammatory Bowel Disease and Type 2 Diabetes in Women consistently demonstrated that MetaGO achieved better prediction accuracy with fewer features compared to previous studies. The experiments showed that MetaGO is a powerful tool for identifying group-specific k-mers, which would be clinically applicable for disease prediction. MetaGO is available at https://github.com/VVsmileyx/MetaGO. PMID:29774017

Fungal diversity in deep-sea sediments of a hydrothermal vent system in the Southwest Indian Ridge

NASA Astrophysics Data System (ADS)

Xu, Wei; Gong, Lin-feng; Pang, Ka-Lai; Luo, Zhu-Hua

2018-01-01

Deep-sea hydrothermal sediment is known to support remarkably diverse microbial consortia. In deep sea environments, fungal communities remain less studied despite their known taxonomic and functional diversity. High-throughput sequencing methods have augmented our capacity to assess eukaryotic diversity and their functions in microbial ecology. Here we provide the first description of the fungal community diversity found in deep sea sediments collected at the Southwest Indian Ridge (SWIR) using culture-dependent and high-throughput sequencing approaches. A total of 138 fungal isolates were cultured from seven different sediment samples using various nutrient media, and these isolates were identified to 14 fungal taxa, including 11 Ascomycota taxa (7 genera) and 3 Basidiomycota taxa (2 genera) based on internal transcribed spacers (ITS1, ITS2 and 5.8S) of rDNA. Using illumina HiSeq sequencing, a total of 757,467 fungal ITS2 tags were recovered from the samples and clustered into 723 operational taxonomic units (OTUs) belonging to 79 taxa (Ascomycota and Basidiomycota contributed to 99% of all samples) based on 97% sequence similarity. Results from both approaches suggest that there is a high fungal diversity in the deep-sea sediments collected in the SWIR and fungal communities were shown to be slightly different by location, although all were collected from adjacent sites at the SWIR. This study provides baseline data of the fungal diversity and biogeography, and a glimpse to the microbial ecology associated with the deep-sea sediments of the hydrothermal vent system of the Southwest Indian Ridge.

Evolutionary relationships of the old world fruit bats (Chiroptera, Pteropodidae): Another star phylogeny?

PubMed Central

2011-01-01

Background The family Pteropodidae comprises bats commonly known as megabats or Old World fruit bats. Molecular phylogenetic studies of pteropodids have provided considerable insight into intrafamilial relationships, but these studies have included only a fraction of the extant diversity (a maximum of 26 out of the 46 currently recognized genera) and have failed to resolve deep relationships among internal clades. Here we readdress the systematics of pteropodids by applying a strategy to try to resolve ancient relationships within Pteropodidae, while providing further insight into subgroup membership, by 1) increasing the taxonomic sample to 42 genera; 2) increasing the number of characters (to >8,000 bp) and nuclear genomic representation; 3) minimizing missing data; 4) controlling for sequence bias; and 5) using appropriate data partitioning and models of sequence evolution. Results Our analyses recovered six principal clades and one additional independent lineage (consisting of a single genus) within Pteropodidae. Reciprocal monophyly of these groups was highly supported and generally congruent among the different methods and datasets used. Likewise, most relationships within these principal clades were well resolved and statistically supported. Relationships among the 7 principal groups, however, were poorly supported in all analyses. This result could not be explained by any detectable systematic bias in the data or incongruence among loci. The SOWH test confirmed that basal branches' lengths were not different from zero, which points to closely-spaced cladogenesis as the most likely explanation for the poor resolution of the deep pteropodid relationships. Simulations suggest that an increase in the amount of sequence data is likely to solve this problem. Conclusions The phylogenetic hypothesis generated here provides a robust framework for a revised cladistic classification of Pteropodidae into subfamilies and tribes and will greatly contribute to the understanding of character evolution and biogeography of pteropodids. The inability of our data to resolve the deepest relationships of the major pteropodid lineages suggests an explosive diversification soon after origin of the crown pteropodids. Several characteristics of pteropodids are consistent with this conclusion, including high species diversity, great morphological diversity, and presence of key innovations in relation to their sister group. PMID:21961908

Data augmentation-assisted deep learning of hand-drawn partially colored sketches for visual search

PubMed Central

Muhammad, Khan; Baik, Sung Wook

2017-01-01

In recent years, image databases are growing at exponential rates, making their management, indexing, and retrieval, very challenging. Typical image retrieval systems rely on sample images as queries. However, in the absence of sample query images, hand-drawn sketches are also used. The recent adoption of touch screen input devices makes it very convenient to quickly draw shaded sketches of objects to be used for querying image databases. This paper presents a mechanism to provide access to visual information based on users’ hand-drawn partially colored sketches using touch screen devices. A key challenge for sketch-based image retrieval systems is to cope with the inherent ambiguity in sketches due to the lack of colors, textures, shading, and drawing imperfections. To cope with these issues, we propose to fine-tune a deep convolutional neural network (CNN) using augmented dataset to extract features from partially colored hand-drawn sketches for query specification in a sketch-based image retrieval framework. The large augmented dataset contains natural images, edge maps, hand-drawn sketches, de-colorized, and de-texturized images which allow CNN to effectively model visual contents presented to it in a variety of forms. The deep features extracted from CNN allow retrieval of images using both sketches and full color images as queries. We also evaluated the role of partial coloring or shading in sketches to improve the retrieval performance. The proposed method is tested on two large datasets for sketch recognition and sketch-based image retrieval and achieved better classification and retrieval performance than many existing methods. PMID:28859140

Comparing species tree estimation with large anchored phylogenomic and small Sanger-sequenced molecular datasets: an empirical study on Malagasy pseudoxyrhophiine snakes.

PubMed

Ruane, Sara; Raxworthy, Christopher J; Lemmon, Alan R; Lemmon, Emily Moriarty; Burbrink, Frank T

2015-10-12

Using molecular data generated by high throughput next generation sequencing (NGS) platforms to infer phylogeny is becoming common as costs go down and the ability to capture loci from across the genome goes up. While there is a general consensus that greater numbers of independent loci should result in more robust phylogenetic estimates, few studies have compared phylogenies resulting from smaller datasets for commonly used genetic markers with the large datasets captured using NGS. Here, we determine how a 5-locus Sanger dataset compares with a 377-locus anchored genomics dataset for understanding the evolutionary history of the pseudoxyrhophiine snake radiation centered in Madagascar. The Pseudoxyrhophiinae comprise ~86 % of Madagascar's serpent diversity, yet they are poorly known with respect to ecology, behavior, and systematics. Using the 377-locus NGS dataset and the summary statistics species-tree methods STAR and MP-EST, we estimated a well-supported species tree that provides new insights concerning intergeneric relationships for the pseudoxyrhophiines. We also compared how these and other methods performed with respect to estimating tree topology using datasets with varying numbers of loci. Using Sanger sequencing and an anchored phylogenomics approach, we sequenced datasets comprised of 5 and 377 loci, respectively, for 23 pseudoxyrhophiine taxa. For each dataset, we estimated phylogenies using both gene-tree (concatenation) and species-tree (STAR, MP-EST) approaches. We determined the similarity of resulting tree topologies from the different datasets using Robinson-Foulds distances. In addition, we examined how subsets of these data performed compared to the complete Sanger and anchored datasets for phylogenetic accuracy using the same tree inference methodologies, as well as the program *BEAST to determine if a full coalescent model for species tree estimation could generate robust results with fewer loci compared to the summary statistics species tree approaches. We also examined the individual gene trees in comparison to the 377-locus species tree using the program MetaTree. Using the full anchored dataset under a variety of methods gave us the same, well-supported phylogeny for pseudoxyrhophiines. The African pseudoxyrhophiine Duberria is the sister taxon to the Malagasy pseudoxyrhophiines genera, providing evidence for a monophyletic radiation in Madagascar. In addition, within Madagascar, the two major clades inferred correspond largely to the aglyphous and opisthoglyphous genera, suggesting that feeding specializations associated with tooth venom delivery may have played a major role in the early diversification of this radiation. The comparison of tree topologies from the concatenated and species-tree methods using different datasets indicated the 5-locus dataset cannot beused to infer a correct phylogeny for the pseudoxyrhophiines under any method tested here and that summary statistics methods require 50 or more loci to consistently recover the species-tree inferred using the complete anchored dataset. However, as few as 15 loci may infer the correct topology when using the full coalescent species tree method *BEAST. MetaTree analyses of each gene tree from the Sanger and anchored datasets found that none of the individual gene trees matched the 377-locus species tree, and that no gene trees were identical with respect to topology. Our results suggest that ≥50 loci may be necessary to confidently infer phylogenies when using summaryspecies-tree methods, but that the coalescent-based method *BEAST consistently recovers the same topology using only 15 loci. These results reinforce that datasets with small numbers of markers may result in misleading topologies, and further, that the method of inference used to generate a phylogeny also has a major influence on the number of loci necessary to infer robust species trees.

Adaptive template generation for amyloid PET using a deep learning approach.

PubMed

Kang, Seung Kwan; Seo, Seongho; Shin, Seong A; Byun, Min Soo; Lee, Dong Young; Kim, Yu Kyeong; Lee, Dong Soo; Lee, Jae Sung

2018-05-11

Accurate spatial normalization (SN) of amyloid positron emission tomography (PET) images for Alzheimer's disease assessment without coregistered anatomical magnetic resonance imaging (MRI) of the same individual is technically challenging. In this study, we applied deep neural networks to generate individually adaptive PET templates for robust and accurate SN of amyloid PET without using matched 3D MR images. Using 681 pairs of simultaneously acquired 11 C-PIB PET and T1-weighted 3D MRI scans of AD, MCI, and cognitively normal subjects, we trained and tested two deep neural networks [convolutional auto-encoder (CAE) and generative adversarial network (GAN)] that produce adaptive best PET templates. More specifically, the networks were trained using 685,100 pieces of augmented data generated by rotating 527 randomly selected datasets and validated using 154 datasets. The input to the supervised neural networks was the 3D PET volume in native space and the label was the spatially normalized 3D PET image using the transformation parameters obtained from MRI-based SN. The proposed deep learning approach significantly enhanced the quantitative accuracy of MRI-less amyloid PET assessment by reducing the SN error observed when an average amyloid PET template is used. Given an input image, the trained deep neural networks rapidly provide individually adaptive 3D PET templates without any discontinuity between the slices (in 0.02 s). As the proposed method does not require 3D MRI for the SN of PET images, it has great potential for use in routine analysis of amyloid PET images in clinical practice and research. © 2018 Wiley Periodicals, Inc.

HLA Diversity in the 1000 Genomes Dataset

PubMed Central

Gourraud, Pierre-Antoine; Khankhanian, Pouya; Cereb, Nezih; Yang, Soo Young; Feolo, Michael; Maiers, Martin; D. Rioux, John; Hauser, Stephen; Oksenberg, Jorge

2014-01-01

The 1000 Genomes Project aims to provide a deep characterization of human genome sequence variation by sequencing at a level that should allow the genome-wide detection of most variants with frequencies as low as 1%. However, in the major histocompatibility complex (MHC), only the top 10 most frequent haplotypes are in the 1% frequency range whereas thousands of haplotypes are present at lower frequencies. Given the limitation of both the coverage and the read length of the sequences generated by the 1000 Genomes Project, the highly variable positions that define HLA alleles may be difficult to identify. We used classical Sanger sequencing techniques to type the HLA-A, HLA-B, HLA-C, HLA-DRB1 and HLA-DQB1 genes in the available 1000 Genomes samples and combined the results with the 103,310 variants in the MHC region genotyped by the 1000 Genomes Project. Using pairwise identity-by-descent distances between individuals and principal component analysis, we established the relationship between ancestry and genetic diversity in the MHC region. As expected, both the MHC variants and the HLA phenotype can identify the major ancestry lineage, informed mainly by the most frequent HLA haplotypes. To some extent, regions of the genome with similar genetic or similar recombination rate have similar properties. An MHC-centric analysis underlines departures between the ancestral background of the MHC and the genome-wide picture. Our analysis of linkage disequilibrium (LD) decay in these samples suggests that overestimation of pairwise LD occurs due to a limited sampling of the MHC diversity. This collection of HLA-specific MHC variants, available on the dbMHC portal, is a valuable resource for future analyses of the role of MHC in population and disease studies. PMID:24988075

HLA diversity in the 1000 genomes dataset.

PubMed

Gourraud, Pierre-Antoine; Khankhanian, Pouya; Cereb, Nezih; Yang, Soo Young; Feolo, Michael; Maiers, Martin; Rioux, John D; Hauser, Stephen; Oksenberg, Jorge

2014-01-01

The 1000 Genomes Project aims to provide a deep characterization of human genome sequence variation by sequencing at a level that should allow the genome-wide detection of most variants with frequencies as low as 1%. However, in the major histocompatibility complex (MHC), only the top 10 most frequent haplotypes are in the 1% frequency range whereas thousands of haplotypes are present at lower frequencies. Given the limitation of both the coverage and the read length of the sequences generated by the 1000 Genomes Project, the highly variable positions that define HLA alleles may be difficult to identify. We used classical Sanger sequencing techniques to type the HLA-A, HLA-B, HLA-C, HLA-DRB1 and HLA-DQB1 genes in the available 1000 Genomes samples and combined the results with the 103,310 variants in the MHC region genotyped by the 1000 Genomes Project. Using pairwise identity-by-descent distances between individuals and principal component analysis, we established the relationship between ancestry and genetic diversity in the MHC region. As expected, both the MHC variants and the HLA phenotype can identify the major ancestry lineage, informed mainly by the most frequent HLA haplotypes. To some extent, regions of the genome with similar genetic or similar recombination rate have similar properties. An MHC-centric analysis underlines departures between the ancestral background of the MHC and the genome-wide picture. Our analysis of linkage disequilibrium (LD) decay in these samples suggests that overestimation of pairwise LD occurs due to a limited sampling of the MHC diversity. This collection of HLA-specific MHC variants, available on the dbMHC portal, is a valuable resource for future analyses of the role of MHC in population and disease studies.

SeqPig: simple and scalable scripting for large sequencing data sets in Hadoop.

PubMed

Schumacher, André; Pireddu, Luca; Niemenmaa, Matti; Kallio, Aleksi; Korpelainen, Eija; Zanetti, Gianluigi; Heljanko, Keijo

2014-01-01

Hadoop MapReduce-based approaches have become increasingly popular due to their scalability in processing large sequencing datasets. However, as these methods typically require in-depth expertise in Hadoop and Java, they are still out of reach of many bioinformaticians. To solve this problem, we have created SeqPig, a library and a collection of tools to manipulate, analyze and query sequencing datasets in a scalable and simple manner. SeqPigscripts use the Hadoop-based distributed scripting engine Apache Pig, which automatically parallelizes and distributes data processing tasks. We demonstrate SeqPig's scalability over many computing nodes and illustrate its use with example scripts. Available under the open source MIT license at http://sourceforge.net/projects/seqpig/

Bellerophon: a program to detect chimeric sequences in multiple sequence alignments.

PubMed

Huber, Thomas; Faulkner, Geoffrey; Hugenholtz, Philip

2004-09-22

Bellerophon is a program for detecting chimeric sequences in multiple sequence datasets by an adaption of partial treeing analysis. Bellerophon was specifically developed to detect 16S rRNA gene chimeras in PCR-clone libraries of environmental samples but can be applied to other nucleotide sequence alignments. Bellerophon is available as an interactive web server at http://foo.maths.uq.edu.au/~huber/bellerophon.pl

High-throughput sequencing and analysis of the gill tissue transcriptome from the deep-sea hydrothermal vent mussel Bathymodiolus azoricus

PubMed Central

2010-01-01

Background Bathymodiolus azoricus is a deep-sea hydrothermal vent mussel found in association with large faunal communities living in chemosynthetic environments at the bottom of the sea floor near the Azores Islands. Investigation of the exceptional physiological reactions that vent mussels have adopted in their habitat, including responses to environmental microbes, remains a difficult challenge for deep-sea biologists. In an attempt to reveal genes potentially involved in the deep-sea mussel innate immunity we carried out a high-throughput sequence analysis of freshly collected B. azoricus transcriptome using gills tissues as the primary source of immune transcripts given its strategic role in filtering the surrounding waterborne potentially infectious microorganisms. Additionally, a substantial EST data set was produced and from which a comprehensive collection of genes coding for putative proteins was organized in a dedicated database, "DeepSeaVent" the first deep-sea vent animal transcriptome database based on the 454 pyrosequencing technology. Results A normalized cDNA library from gills tissue was sequenced in a full 454 GS-FLX run, producing 778,996 sequencing reads. Assembly of the high quality reads resulted in 75,407 contigs of which 3,071 were singletons. A total of 39,425 transcripts were conceptually translated into amino-sequences of which 22,023 matched known proteins in the NCBI non-redundant protein database, 15,839 revealed conserved protein domains through InterPro functional classification and 9,584 were assigned with Gene Ontology terms. Queries conducted within the database enabled the identification of genes putatively involved in immune and inflammatory reactions which had not been previously evidenced in the vent mussel. Their physical counterpart was confirmed by semi-quantitative quantitative Reverse-Transcription-Polymerase Chain Reactions (RT-PCR) and their RNA transcription level by quantitative PCR (qPCR) experiments. Conclusions We have established the first tissue transcriptional analysis of a deep-sea hydrothermal vent animal and generated a searchable catalog of genes that provides a direct method of identifying and retrieving vast numbers of novel coding sequences which can be applied in gene expression profiling experiments from a non-conventional model organism. This provides the most comprehensive sequence resource for identifying novel genes currently available for a deep-sea vent organism, in particular, genes putatively involved in immune and inflammatory reactions in vent mussels. The characterization of the B. azoricus transcriptome will facilitate research into biological processes underlying physiological adaptations to hydrothermal vent environments and will provide a basis for expanding our understanding of genes putatively involved in adaptations processes during post-capture long term acclimatization experiments, at "sea-level" conditions, using B. azoricus as a model organism. PMID:20937131

TERATOLOGY v2.0 – building a path forward

EPA Science Inventory

Unraveling the complex relationships between environmental factors and early life susceptibility in assessing the risk for adverse pregnancy outcomes requires advanced knowledge of biological systems. Large datasets and deep data-mining tools are emerging resources for predictive...

DeepGO: predicting protein functions from sequence and interactions using a deep ontology-aware classifier.

PubMed

Kulmanov, Maxat; Khan, Mohammed Asif; Hoehndorf, Robert; Wren, Jonathan

2018-02-15

A large number of protein sequences are becoming available through the application of novel high-throughput sequencing technologies. Experimental functional characterization of these proteins is time-consuming and expensive, and is often only done rigorously for few selected model organisms. Computational function prediction approaches have been suggested to fill this gap. The functions of proteins are classified using the Gene Ontology (GO), which contains over 40 000 classes. Additionally, proteins have multiple functions, making function prediction a large-scale, multi-class, multi-label problem. We have developed a novel method to predict protein function from sequence. We use deep learning to learn features from protein sequences as well as a cross-species protein-protein interaction network. Our approach specifically outputs information in the structure of the GO and utilizes the dependencies between GO classes as background information to construct a deep learning model. We evaluate our method using the standards established by the Computational Assessment of Function Annotation (CAFA) and demonstrate a significant improvement over baseline methods such as BLAST, in particular for predicting cellular locations. Web server: http://deepgo.bio2vec.net, Source code: https://github.com/bio-ontology-research-group/deepgo. robert.hoehndorf@kaust.edu.sa. Supplementary data are available at Bioinformatics online. © The Author(s) 2017. Published by Oxford University Press.

DeepSAT: A Deep Learning Approach to Tree-Cover Delineation in 1-m NAIP Imagery for the Continental United States

NASA Technical Reports Server (NTRS)

Ganguly, Sangram; Basu, Saikat; Nemani, Ramakrishna R.; Mukhopadhyay, Supratik; Michaelis, Andrew; Votava, Petr

2016-01-01

High resolution tree cover classification maps are needed to increase the accuracy of current land ecosystem and climate model outputs. Limited studies are in place that demonstrates the state-of-the-art in deriving very high resolution (VHR) tree cover products. In addition, most methods heavily rely on commercial softwares that are difficult to scale given the region of study (e.g. continents to globe). Complexities in present approaches relate to (a) scalability of the algorithm, (b) large image data processing (compute and memory intensive), (c) computational cost, (d) massively parallel architecture, and (e) machine learning automation. In addition, VHR satellite datasets are of the order of terabytes and features extracted from these datasets are of the order of petabytes. In our present study, we have acquired the National Agriculture Imagery Program (NAIP) dataset for the Continental United States at a spatial resolution of 1-m. This data comes as image tiles (a total of quarter million image scenes with 60 million pixels) and has a total size of 65 terabytes for a single acquisition. Features extracted from the entire dataset would amount to 8-10 petabytes. In our proposed approach, we have implemented a novel semi-automated machine learning algorithm rooted on the principles of "deep learning" to delineate the percentage of tree cover. Using the NASA Earth Exchange (NEX) initiative, we have developed an end-to-end architecture by integrating a segmentation module based on Statistical Region Merging, a classification algorithm using Deep Belief Network and a structured prediction algorithm using Conditional Random Fields to integrate the results from the segmentation and classification modules to create per-pixel class labels. The training process is scaled up using the power of GPUs and the prediction is scaled to quarter million NAIP tiles spanning the whole of Continental United States using the NEX HPC supercomputing cluster. An initial pilot over the state of California spanning a total of 11,095 NAIP tiles covering a total geographical area of 163,696 sq. miles has produced true positive rates of around 88 percent for fragmented forests and 74 percent for urban tree cover areas, with false positive rates lower than 2 percent for both landscapes.

DeepSAT: A Deep Learning Approach to Tree-cover Delineation in 1-m NAIP Imagery for the Continental United States

NASA Astrophysics Data System (ADS)

Ganguly, S.; Basu, S.; Nemani, R. R.; Mukhopadhyay, S.; Michaelis, A.; Votava, P.

2016-12-01

High resolution tree cover classification maps are needed to increase the accuracy of current land ecosystem and climate model outputs. Limited studies are in place that demonstrates the state-of-the-art in deriving very high resolution (VHR) tree cover products. In addition, most methods heavily rely on commercial softwares that are difficult to scale given the region of study (e.g. continents to globe). Complexities in present approaches relate to (a) scalability of the algorithm, (b) large image data processing (compute and memory intensive), (c) computational cost, (d) massively parallel architecture, and (e) machine learning automation. In addition, VHR satellite datasets are of the order of terabytes and features extracted from these datasets are of the order of petabytes. In our present study, we have acquired the National Agriculture Imagery Program (NAIP) dataset for the Continental United States at a spatial resolution of 1-m. This data comes as image tiles (a total of quarter million image scenes with 60 million pixels) and has a total size of 65 terabytes for a single acquisition. Features extracted from the entire dataset would amount to 8-10 petabytes. In our proposed approach, we have implemented a novel semi-automated machine learning algorithm rooted on the principles of "deep learning" to delineate the percentage of tree cover. Using the NASA Earth Exchange (NEX) initiative, we have developed an end-to-end architecture by integrating a segmentation module based on Statistical Region Merging, a classification algorithm using Deep Belief Network and a structured prediction algorithm using Conditional Random Fields to integrate the results from the segmentation and classification modules to create per-pixel class labels. The training process is scaled up using the power of GPUs and the prediction is scaled to quarter million NAIP tiles spanning the whole of Continental United States using the NEX HPC supercomputing cluster. An initial pilot over the state of California spanning a total of 11,095 NAIP tiles covering a total geographical area of 163,696 sq. miles has produced true positive rates of around 88% for fragmented forests and 74% for urban tree cover areas, with false positive rates lower than 2% for both landscapes.

Using Digital Globes to Explore the Deep Sea and Advance Public Literacy in Earth System Science

NASA Astrophysics Data System (ADS)

Beaulieu, S. E.; Brickley, A.; Emery, M.; Spargo, A.; Patterson, K.; Joyce, K.; Silva, T.; Madin, K.

2014-12-01

Digital globes are new technologies increasingly used in both informal and formal education to display global datasets. By creating a narrative using multiple datasets, linkages between Earth systems - lithosphere, hydrosphere, atmosphere, and biosphere - can be conveyed. But how effective are digital globes in advancing public literacy in Earth system science? We addressed this question in developing new content for digital globes that interweaves imagery obtained by deep-diving vehicles with global datasets, including a new dataset locating the world's known hydrothermal vents. Our two narratives, "Life Without Sunlight" (LWS) and "Smoke and Fire Underwater" (SFU), each focus on STEM (science, technology, engineering, and mathematics) principles related to geology, biology, and exploration. We are preparing a summative evaluation for our content delivered on NOAA's Science on a Sphere as interactive presentations and as movies. We tested knowledge gained with respect to the STEM principles and the level of excitement generated by the virtual deep-sea exploration. We conducted a Post-test Only Design with quantitative data based on self-reporting on a Likert scale. A total of 75 adults and 48 youths responded to our questionnaire, distributed into test groups that saw either one of the two narratives delivered either as a movie or as an interactive presentation. Here, we report preliminary results for the youths, the majority (81%) of which live in towns with lower income and lower levels of educational attainment as compared to other towns in Massachusetts. For both narratives, there was knowledge gained for all 6 STEM principles and "Quite a Bit" of excitement. The mode in responses for knowledge gained was "Quite a Bit" for both the movie and the interactive presentation for 4 of the STEM principles (LWS geology, LWS biology, SFU geology, and SFU exploration) and "Some" for SFU biology. Only for LWS exploration was there a difference in mode between the interactive presentation ("A Little") and the movie ("Quite a Bit"). We conclude that our content for digital globes is effective in teaching the STEM principles and exciting viewers about the deep ocean frontier. We attribute this success to the tight collaboration between scientists, educators, and graphic artists in developing the content for public audiences.

Do deep convolutional neural networks really need to be deep when applied for remote scene classification?

NASA Astrophysics Data System (ADS)

Luo, Chang; Wang, Jie; Feng, Gang; Xu, Suhui; Wang, Shiqiang

2017-10-01

Deep convolutional neural networks (CNNs) have been widely used to obtain high-level representation in various computer vision tasks. However, for remote scene classification, there are not sufficient images to train a very deep CNN from scratch. From two viewpoints of generalization power, we propose two promising kinds of deep CNNs for remote scenes and try to find whether deep CNNs need to be deep for remote scene classification. First, we transfer successful pretrained deep CNNs to remote scenes based on the theory that depth of CNNs brings the generalization power by learning available hypothesis for finite data samples. Second, according to the opposite viewpoint that generalization power of deep CNNs comes from massive memorization and shallow CNNs with enough neural nodes have perfect finite sample expressivity, we design a lightweight deep CNN (LDCNN) for remote scene classification. With five well-known pretrained deep CNNs, experimental results on two independent remote-sensing datasets demonstrate that transferred deep CNNs can achieve state-of-the-art results in an unsupervised setting. However, because of its shallow architecture, LDCNN cannot obtain satisfactory performance, regardless of whether in an unsupervised, semisupervised, or supervised setting. CNNs really need depth to obtain general features for remote scenes. This paper also provides baseline for applying deep CNNs to other remote sensing tasks.

Integrative Data Analysis of Multi-Platform Cancer Data with a Multimodal Deep Learning Approach.

PubMed

Liang, Muxuan; Li, Zhizhong; Chen, Ting; Zeng, Jianyang

2015-01-01

Identification of cancer subtypes plays an important role in revealing useful insights into disease pathogenesis and advancing personalized therapy. The recent development of high-throughput sequencing technologies has enabled the rapid collection of multi-platform genomic data (e.g., gene expression, miRNA expression, and DNA methylation) for the same set of tumor samples. Although numerous integrative clustering approaches have been developed to analyze cancer data, few of them are particularly designed to exploit both deep intrinsic statistical properties of each input modality and complex cross-modality correlations among multi-platform input data. In this paper, we propose a new machine learning model, called multimodal deep belief network (DBN), to cluster cancer patients from multi-platform observation data. In our integrative clustering framework, relationships among inherent features of each single modality are first encoded into multiple layers of hidden variables, and then a joint latent model is employed to fuse common features derived from multiple input modalities. A practical learning algorithm, called contrastive divergence (CD), is applied to infer the parameters of our multimodal DBN model in an unsupervised manner. Tests on two available cancer datasets show that our integrative data analysis approach can effectively extract a unified representation of latent features to capture both intra- and cross-modality correlations, and identify meaningful disease subtypes from multi-platform cancer data. In addition, our approach can identify key genes and miRNAs that may play distinct roles in the pathogenesis of different cancer subtypes. Among those key miRNAs, we found that the expression level of miR-29a is highly correlated with survival time in ovarian cancer patients. These results indicate that our multimodal DBN based data analysis approach may have practical applications in cancer pathogenesis studies and provide useful guidelines for personalized cancer therapy.

Contourite drifts on early passive margins as an indicator of established lithospheric breakup

NASA Astrophysics Data System (ADS)

Soares, Duarte M.; Alves, Tiago M.; Terrinha, Pedro

2014-09-01

The Albian-Cenomanian breakup sequence (BS) offshore Northwest Iberia is mapped, described and characterised for the first time in terms of its seismic and depositional facies. The interpreted dataset comprises a large grid of regional (2D) seismic-reflection profiles, complemented by Industry and ODP/DSDP borehole data. Within the BS are observed distinct seismic facies that reflect the presence of: (a) black shales and fine-grained turbidites, (b) mass-transport deposits (MTDs) and coarse-grained turbidites, and (c) contourite drifts. Borehole data show that these depositional systems developed as mixed carbonate-siliciclastic sediments proximally, and as organic-carbon-rich mudstones (black shales) distally on the Northwest Iberia margin. MTDs and turbidites tend to occur on the continental slope, frequently in association with large-scale olistostromes. Distally, these change into interbedded fine-grained turbidites and black shales showing widespread evidence of deep-water current activity towards the top of the BS. Current activity is expressed by intra-BS erosional surfaces and sediment drifts. The results in this paper are important as they demonstrate that contourite drifts are ubiquitous features in the study area after Aptian-Albian lithospheric breakup. Therefore, we interpret the recognition of contourite drifts in Northwest Iberia as having significant palaeogeographic implications. Contourite drifts materialise the onset of important deep-water circulation marking the establishment of oceanic gateways between two fully separated continental margins. As a corollary, we postulate the generation of deep-water geostrophic currents to have had significant impact on North Atlantic climate and ocean circulation during the Albian-Cenomanian, with the record of such impacts being preserved in the contourite drifts analysed in this work.

A deep learning framework for modeling structural features of RNA-binding protein targets

PubMed Central

Zhang, Sai; Zhou, Jingtian; Hu, Hailin; Gong, Haipeng; Chen, Ligong; Cheng, Chao; Zeng, Jianyang

2016-01-01

RNA-binding proteins (RBPs) play important roles in the post-transcriptional control of RNAs. Identifying RBP binding sites and characterizing RBP binding preferences are key steps toward understanding the basic mechanisms of the post-transcriptional gene regulation. Though numerous computational methods have been developed for modeling RBP binding preferences, discovering a complete structural representation of the RBP targets by integrating their available structural features in all three dimensions is still a challenging task. In this paper, we develop a general and flexible deep learning framework for modeling structural binding preferences and predicting binding sites of RBPs, which takes (predicted) RNA tertiary structural information into account for the first time. Our framework constructs a unified representation that characterizes the structural specificities of RBP targets in all three dimensions, which can be further used to predict novel candidate binding sites and discover potential binding motifs. Through testing on the real CLIP-seq datasets, we have demonstrated that our deep learning framework can automatically extract effective hidden structural features from the encoded raw sequence and structural profiles, and predict accurate RBP binding sites. In addition, we have conducted the first study to show that integrating the additional RNA tertiary structural features can improve the model performance in predicting RBP binding sites, especially for the polypyrimidine tract-binding protein (PTB), which also provides a new evidence to support the view that RBPs may own specific tertiary structural binding preferences. In particular, the tests on the internal ribosome entry site (IRES) segments yield satisfiable results with experimental support from the literature and further demonstrate the necessity of incorporating RNA tertiary structural information into the prediction model. The source code of our approach can be found in https://github.com/thucombio/deepnet-rbp. PMID:26467480

GRAM-CNN: a deep learning approach with local context for named entity recognition in biomedical text.

PubMed

Zhu, Qile; Li, Xiaolin; Conesa, Ana; Pereira, Cécile

2018-05-01

Best performing named entity recognition (NER) methods for biomedical literature are based on hand-crafted features or task-specific rules, which are costly to produce and difficult to generalize to other corpora. End-to-end neural networks achieve state-of-the-art performance without hand-crafted features and task-specific knowledge in non-biomedical NER tasks. However, in the biomedical domain, using the same architecture does not yield competitive performance compared with conventional machine learning models. We propose a novel end-to-end deep learning approach for biomedical NER tasks that leverages the local contexts based on n-gram character and word embeddings via Convolutional Neural Network (CNN). We call this approach GRAM-CNN. To automatically label a word, this method uses the local information around a word. Therefore, the GRAM-CNN method does not require any specific knowledge or feature engineering and can be theoretically applied to a wide range of existing NER problems. The GRAM-CNN approach was evaluated on three well-known biomedical datasets containing different BioNER entities. It obtained an F1-score of 87.26% on the Biocreative II dataset, 87.26% on the NCBI dataset and 72.57% on the JNLPBA dataset. Those results put GRAM-CNN in the lead of the biological NER methods. To the best of our knowledge, we are the first to apply CNN based structures to BioNER problems. The GRAM-CNN source code, datasets and pre-trained model are available online at: https://github.com/valdersoul/GRAM-CNN. andyli@ece.ufl.edu or aconesa@ufl.edu. Supplementary data are available at Bioinformatics online.

GRAM-CNN: a deep learning approach with local context for named entity recognition in biomedical text

PubMed Central

Zhu, Qile; Li, Xiaolin; Conesa, Ana; Pereira, Cécile

2018-01-01

Abstract Motivation Best performing named entity recognition (NER) methods for biomedical literature are based on hand-crafted features or task-specific rules, which are costly to produce and difficult to generalize to other corpora. End-to-end neural networks achieve state-of-the-art performance without hand-crafted features and task-specific knowledge in non-biomedical NER tasks. However, in the biomedical domain, using the same architecture does not yield competitive performance compared with conventional machine learning models. Results We propose a novel end-to-end deep learning approach for biomedical NER tasks that leverages the local contexts based on n-gram character and word embeddings via Convolutional Neural Network (CNN). We call this approach GRAM-CNN. To automatically label a word, this method uses the local information around a word. Therefore, the GRAM-CNN method does not require any specific knowledge or feature engineering and can be theoretically applied to a wide range of existing NER problems. The GRAM-CNN approach was evaluated on three well-known biomedical datasets containing different BioNER entities. It obtained an F1-score of 87.26% on the Biocreative II dataset, 87.26% on the NCBI dataset and 72.57% on the JNLPBA dataset. Those results put GRAM-CNN in the lead of the biological NER methods. To the best of our knowledge, we are the first to apply CNN based structures to BioNER problems. Availability and implementation The GRAM-CNN source code, datasets and pre-trained model are available online at: https://github.com/valdersoul/GRAM-CNN. Contact andyli@ece.ufl.edu or aconesa@ufl.edu Supplementary information Supplementary data are available at Bioinformatics online. PMID:29272325

Rescuing Paleomagnetic Data from Deep-Sea Cores Through the IEDA-CCNY Data Internship Program

NASA Astrophysics Data System (ADS)

Ismail, A.; Randel, C.; Palumbo, R. V.; Carter, M.; Cai, Y.; Kent, D. V.; Lehnert, K.; Block, K. A.

2016-12-01

Paleomagnetic data provides essential information for evaluating the chronostratigraphy of sedimentary cores. Lamont research vessels Vema and Robert Conrad collected over 10,000 deep-sea sediment cores around the world from 1953 to 1989. 10% of these cores have been sampled for paleomagnetic analyses at Lamont. Over the years, only 10% of these paleomagnetic records have been published. Moreover, data listings were only rarely made available in older publications because electronic appendices were not available and cyberinfrastructure was not in place for publishing and preserving these data. As a result, the majority of these datasets exist only as fading computer printouts in binders on the investigator's bookshelf. This summer, undergraduate students from the NSF-funded IEDA-CCNY Data Internship Program started digitizing this enormous dataset under the supervision of Dennis Kent, the current custodian of the data and one of the investigators who oversaw some of the data collection process, and an active leader in the field. Undergraduate students worked on digitizing paper records, proof-reading and organizing the data sheets for future integration into an appropriate repository. Through observing and plotting the data, the students learned about how sediment cores and paleomagnetic data are collected and used in research, and the best practices in data publishing and preservation from IEDA (Interdisciplinary Earth Data Alliance) team members. The students also compared different optical character recognition (OCR) softwares and established an efficient workflow to digitize these datasets. These datasets will eventually be incorporated in the Magnetics Information Consortium (MagIC), so that they can be easily compared with similar datasets and have the potential to generate new findings. Through this data rescue project, the students had the opportunity to learn about an important field of scientific research and interact with world-class scientists.

Efficacy of a Deep Learning System for Detecting Glaucomatous Optic Neuropathy Based on Color Fundus Photographs.

PubMed

Li, Zhixi; He, Yifan; Keel, Stuart; Meng, Wei; Chang, Robert T; He, Mingguang

2018-03-02

To assess the performance of a deep learning algorithm for detecting referable glaucomatous optic neuropathy (GON) based on color fundus photographs. A deep learning system for the classification of GON was developed for automated classification of GON on color fundus photographs. We retrospectively included 48 116 fundus photographs for the development and validation of a deep learning algorithm. This study recruited 21 trained ophthalmologists to classify the photographs. Referable GON was defined as vertical cup-to-disc ratio of 0.7 or more and other typical changes of GON. The reference standard was made until 3 graders achieved agreement. A separate validation dataset of 8000 fully gradable fundus photographs was used to assess the performance of this algorithm. The area under receiver operator characteristic curve (AUC) with sensitivity and specificity was applied to evaluate the efficacy of the deep learning algorithm detecting referable GON. In the validation dataset, this deep learning system achieved an AUC of 0.986 with sensitivity of 95.6% and specificity of 92.0%. The most common reasons for false-negative grading (n = 87) were GON with coexisting eye conditions (n = 44 [50.6%]), including pathologic or high myopia (n = 37 [42.6%]), diabetic retinopathy (n = 4 [4.6%]), and age-related macular degeneration (n = 3 [3.4%]). The leading reason for false-positive results (n = 480) was having other eye conditions (n = 458 [95.4%]), mainly including physiologic cupping (n = 267 [55.6%]). Misclassification as false-positive results amidst a normal-appearing fundus occurred in only 22 eyes (4.6%). A deep learning system can detect referable GON with high sensitivity and specificity. Coexistence of high or pathologic myopia is the most common cause resulting in false-negative results. Physiologic cupping and pathologic myopia were the most common reasons for false-positive results. Copyright © 2018 American Academy of Ophthalmology. Published by Elsevier Inc. All rights reserved.

Sensitive Deep-Sequencing-Based HIV-1 Genotyping Assay To Simultaneously Determine Susceptibility to Protease, Reverse Transcriptase, Integrase, and Maturation Inhibitors, as Well as HIV-1 Coreceptor Tropism

PubMed Central

Gibson, Richard M.; Meyer, Ashley M.; Winner, Dane; Archer, John; Feyertag, Felix; Ruiz-Mateos, Ezequiel; Leal, Manuel; Robertson, David L.; Schmotzer, Christine L.

2014-01-01

With 29 individual antiretroviral drugs available from six classes that are approved for the treatment of HIV-1 infection, a combination of different phenotypic and genotypic tests is currently needed to monitor HIV-infected individuals. In this study, we developed a novel HIV-1 genotypic assay based on deep sequencing (DeepGen HIV) to simultaneously assess HIV-1 susceptibilities to all drugs targeting the three viral enzymes and to predict HIV-1 coreceptor tropism. Patient-derived gag-p2/NCp7/p1/p6/pol-PR/RT/IN- and env-C2V3 PCR products were sequenced using the Ion Torrent Personal Genome Machine. Reads spanning the 3′ end of the Gag, protease (PR), reverse transcriptase (RT), integrase (IN), and V3 regions were extracted, truncated, translated, and assembled for genotype and HIV-1 coreceptor tropism determination. DeepGen HIV consistently detected both minority drug-resistant viruses and non-R5 HIV-1 variants from clinical specimens with viral loads of ≥1,000 copies/ml and from B and non-B subtypes. Additional mutations associated with resistance to PR, RT, and IN inhibitors, previously undetected by standard (Sanger) population sequencing, were reliably identified at frequencies as low as 1%. DeepGen HIV results correlated with phenotypic (original Trofile, 92%; enhanced-sensitivity Trofile assay [ESTA], 80%; TROCAI, 81%; and VeriTrop, 80%) and genotypic (population sequencing/Geno2Pheno with a 10% false-positive rate [FPR], 84%) HIV-1 tropism test results. DeepGen HIV (83%) and Trofile (85%) showed similar concordances with the clinical response following an 8-day course of maraviroc monotherapy (MCT). In summary, this novel all-inclusive HIV-1 genotypic and coreceptor tropism assay, based on deep sequencing of the PR, RT, IN, and V3 regions, permits simultaneous multiplex detection of low-level drug-resistant and/or non-R5 viruses in up to 96 clinical samples. This comprehensive test, the first of its class, will be instrumental in the development of new antiretroviral drugs and, more importantly, will aid in the treatment and management of HIV-infected individuals. PMID:24468782

Dendrites, deep learning, and sequences in the hippocampus.

PubMed

Bhalla, Upinder S

2017-10-12

The hippocampus places us both in time and space. It does so over remarkably large spans: milliseconds to years, and centimeters to kilometers. This works for sensory representations, for memory, and for behavioral context. How does it fit in such wide ranges of time and space scales, and keep order among the many dimensions of stimulus context? A key organizing principle for a wide sweep of scales and stimulus dimensions is that of order in time, or sequences. Sequences of neuronal activity are ubiquitous in sensory processing, in motor control, in planning actions, and in memory. Against this strong evidence for the phenomenon, there are currently more models than definite experiments about how the brain generates ordered activity. The flip side of sequence generation is discrimination. Discrimination of sequences has been extensively studied at the behavioral, systems, and modeling level, but again physiological mechanisms are fewer. It is against this backdrop that I discuss two recent developments in neural sequence computation, that at face value share little beyond the label "neural." These are dendritic sequence discrimination, and deep learning. One derives from channel physiology and molecular signaling, the other from applied neural network theory - apparently extreme ends of the spectrum of neural circuit detail. I suggest that each of these topics has deep lessons about the possible mechanisms, scales, and capabilities of hippocampal sequence computation. © 2017 Wiley Periodicals, Inc.

MRUniNovo: an efficient tool for de novo peptide sequencing utilizing the hadoop distributed computing framework.

PubMed

Li, Chuang; Chen, Tao; He, Qiang; Zhu, Yunping; Li, Kenli

2017-03-15

Tandem mass spectrometry-based de novo peptide sequencing is a complex and time-consuming process. The current algorithms for de novo peptide sequencing cannot rapidly and thoroughly process large mass spectrometry datasets. In this paper, we propose MRUniNovo, a novel tool for parallel de novo peptide sequencing. MRUniNovo parallelizes UniNovo based on the Hadoop compute platform. Our experimental results demonstrate that MRUniNovo significantly reduces the computation time of de novo peptide sequencing without sacrificing the correctness and accuracy of the results, and thus can process very large datasets that UniNovo cannot. MRUniNovo is an open source software tool implemented in java. The source code and the parameter settings are available at http://bioinfo.hupo.org.cn/MRUniNovo/index.php. s131020002@hnu.edu.cn ; taochen1019@163.com. Supplementary data are available at Bioinformatics online. © The Author 2016. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com

A robust and cost-effective approach to sequence and analyze complete genomes of small RNA viruses

USDA-ARS?s Scientific Manuscript database

Background: Next-generation sequencing (NGS) allows ultra-deep sequencing of nucleic acids. The use of sequence-independent amplification of viral nucleic acids without utilization of target-specific primers provides advantages over traditional sequencing methods and allows detection of unsuspected ...

De novo transcriptome assembly and positive selection analysis of an individual deep-sea fish.

PubMed

Lan, Yi; Sun, Jin; Xu, Ting; Chen, Chong; Tian, Renmao; Qiu, Jian-Wen; Qian, Pei-Yuan

2018-05-24

High hydrostatic pressure and low temperatures make the deep sea a harsh environment for life forms. Actin organization and microtubules assembly, which are essential for intracellular transport and cell motility, can be disrupted by high hydrostatic pressure. High hydrostatic pressure can also damage DNA. Nucleic acids exposed to low temperatures can form secondary structures that hinder genetic information processing. To study how deep-sea creatures adapt to such a hostile environment, one of the most straightforward ways is to sequence and compare their genes with those of their shallow-water relatives. We captured an individual of the fish species Aldrovandia affinis, which is a typical deep-sea inhabitant, from the Okinawa Trough at a depth of 1550 m using a remotely operated vehicle (ROV). We sequenced its transcriptome and analyzed its molecular adaptation. We obtained 27,633 protein coding sequences using an Illumina platform and compared them with those of several shallow-water fish species. Analysis of 4918 single-copy orthologs identified 138 positively selected genes in A. affinis, including genes involved in microtubule regulation. Particularly, functional domains related to cold shock as well as DNA repair are exposed to positive selection pressure in both deep-sea fish and hadal amphipod. Overall, we have identified a set of positively selected genes related to cytoskeleton structures, DNA repair and genetic information processing, which shed light on molecular adaptation to the deep sea. These results suggest that amino acid substitutions of these positively selected genes may contribute crucially to the adaptation of deep-sea animals. Additionally, we provide a high-quality transcriptome of a deep-sea fish for future deep-sea studies.

Systematic evaluation of deep learning based detection frameworks for aerial imagery

NASA Astrophysics Data System (ADS)

Sommer, Lars; Steinmann, Lucas; Schumann, Arne; Beyerer, Jürgen

2018-04-01

Object detection in aerial imagery is crucial for many applications in the civil and military domain. In recent years, deep learning based object detection frameworks significantly outperformed conventional approaches based on hand-crafted features on several datasets. However, these detection frameworks are generally designed and optimized for common benchmark datasets, which considerably differ from aerial imagery especially in object sizes. As already demonstrated for Faster R-CNN, several adaptations are necessary to account for these differences. In this work, we adapt several state-of-the-art detection frameworks including Faster R-CNN, R-FCN, and Single Shot MultiBox Detector (SSD) to aerial imagery. We discuss adaptations that mainly improve the detection accuracy of all frameworks in detail. As the output of deeper convolutional layers comprise more semantic information, these layers are generally used in detection frameworks as feature map to locate and classify objects. However, the resolution of these feature maps is insufficient for handling small object instances, which results in an inaccurate localization or incorrect classification of small objects. Furthermore, state-of-the-art detection frameworks perform bounding box regression to predict the exact object location. Therefore, so called anchor or default boxes are used as reference. We demonstrate how an appropriate choice of anchor box sizes can considerably improve detection performance. Furthermore, we evaluate the impact of the performed adaptations on two publicly available datasets to account for various ground sampling distances or differing backgrounds. The presented adaptations can be used as guideline for further datasets or detection frameworks.

Revealing the unexplored fungal communities in deep groundwater of crystalline bedrock fracture zones in Olkiluoto, Finland.

PubMed

Sohlberg, Elina; Bomberg, Malin; Miettinen, Hanna; Nyyssönen, Mari; Salavirta, Heikki; Vikman, Minna; Itävaara, Merja

2015-01-01

The diversity and functional role of fungi, one of the ecologically most important groups of eukaryotic microorganisms, remains largely unknown in deep biosphere environments. In this study we investigated fungal communities in packer-isolated bedrock fractures in Olkiluoto, Finland at depths ranging from 296 to 798 m below surface level. DNA- and cDNA-based high-throughput amplicon sequencing analysis of the fungal internal transcribed spacer (ITS) gene markers was used to examine the total fungal diversity and to identify the active members in deep fracture zones at different depths. Results showed that fungi were present in fracture zones at all depths and fungal diversity was higher than expected. Most of the observed fungal sequences belonged to the phylum Ascomycota. Phyla Basidiomycota and Chytridiomycota were only represented as a minor part of the fungal community. Dominating fungal classes in the deep bedrock aquifers were Sordariomycetes, Eurotiomycetes, and Dothideomycetes from the Ascomycota phylum and classes Microbotryomycetes and Tremellomycetes from the Basidiomycota phylum, which are the most frequently detected fungal taxa reported also from deep sea environments. In addition some fungal sequences represented potentially novel fungal species. Active fungi were detected in most of the fracture zones, which proves that fungi are able to maintain cellular activity in these oligotrophic conditions. Possible roles of fungi and their origin in deep bedrock groundwater can only be speculated in the light of current knowledge but some species may be specifically adapted to deep subsurface environment and may play important roles in the utilization and recycling of nutrients and thus sustaining the deep subsurface microbial community.

User Guidelines for the Brassica Database: BRAD.

PubMed

Wang, Xiaobo; Cheng, Feng; Wang, Xiaowu

2016-01-01

The genome sequence of Brassica rapa was first released in 2011. Since then, further Brassica genomes have been sequenced or are undergoing sequencing. It is therefore necessary to develop tools that help users to mine information from genomic data efficiently. This will greatly aid scientific exploration and breeding application, especially for those with low levels of bioinformatic training. Therefore, the Brassica database (BRAD) was built to collect, integrate, illustrate, and visualize Brassica genomic datasets. BRAD provides useful searching and data mining tools, and facilitates the search of gene annotation datasets, syntenic or non-syntenic orthologs, and flanking regions of functional genomic elements. It also includes genome-analysis tools such as BLAST and GBrowse. One of the important aims of BRAD is to build a bridge between Brassica crop genomes with the genome of the model species Arabidopsis thaliana, thus transferring the bulk of A. thaliana gene study information for use with newly sequenced Brassica crops.

Construction of a large collection of small genome variations in French dairy and beef breeds using whole-genome sequences.

PubMed

Boussaha, Mekki; Michot, Pauline; Letaief, Rabia; Hozé, Chris; Fritz, Sébastien; Grohs, Cécile; Esquerré, Diane; Duchesne, Amandine; Philippe, Romain; Blanquet, Véronique; Phocas, Florence; Floriot, Sandrine; Rocha, Dominique; Klopp, Christophe; Capitan, Aurélien; Boichard, Didier

2016-11-15

In recent years, several bovine genome sequencing projects were carried out with the aim of developing genomic tools to improve dairy and beef production efficiency and sustainability. In this study, we describe the first French cattle genome variation dataset obtained by sequencing 274 whole genomes representing several major dairy and beef breeds. This dataset contains over 28 million single nucleotide polymorphisms (SNPs) and small insertions and deletions. Comparisons between sequencing results and SNP array genotypes revealed a very high genotype concordance rate, which indicates the good quality of our data. To our knowledge, this is the first large-scale catalog of small genomic variations in French dairy and beef cattle. This resource will contribute to the study of gene functions and population structure and also help to improve traits through genotype-guided selection.

Recovering complete and draft population genomes from metagenome datasets

DOE PAGES

Sangwan, Naseer; Xia, Fangfang; Gilbert, Jack A.

2016-03-08

Assembly of metagenomic sequence data into microbial genomes is of fundamental value to improving our understanding of microbial ecology and metabolism by elucidating the functional potential of hard-to-culture microorganisms. Here, we provide a synthesis of available methods to bin metagenomic contigs into species-level groups and highlight how genetic diversity, sequencing depth, and coverage influence binning success. Despite the computational cost on application to deeply sequenced complex metagenomes (e.g., soil), covarying patterns of contig coverage across multiple datasets significantly improves the binning process. We also discuss and compare current genome validation methods and reveal how these methods tackle the problem ofmore » chimeric genome bins i.e., sequences from multiple species. Finally, we explore how population genome assembly can be used to uncover biogeographic trends and to characterize the effect of in situ functional constraints on the genome-wide evolution.« less

Recovering complete and draft population genomes from metagenome datasets

DOE Office of Scientific and Technical Information (OSTI.GOV)

Sangwan, Naseer; Xia, Fangfang; Gilbert, Jack A.

Assembly of metagenomic sequence data into microbial genomes is of fundamental value to improving our understanding of microbial ecology and metabolism by elucidating the functional potential of hard-to-culture microorganisms. Here, we provide a synthesis of available methods to bin metagenomic contigs into species-level groups and highlight how genetic diversity, sequencing depth, and coverage influence binning success. Despite the computational cost on application to deeply sequenced complex metagenomes (e.g., soil), covarying patterns of contig coverage across multiple datasets significantly improves the binning process. We also discuss and compare current genome validation methods and reveal how these methods tackle the problem ofmore » chimeric genome bins i.e., sequences from multiple species. Finally, we explore how population genome assembly can be used to uncover biogeographic trends and to characterize the effect of in situ functional constraints on the genome-wide evolution.« less

Ten years of maintaining and expanding a microbial genome and metagenome analysis system.

PubMed

Markowitz, Victor M; Chen, I-Min A; Chu, Ken; Pati, Amrita; Ivanova, Natalia N; Kyrpides, Nikos C

2015-11-01

Launched in March 2005, the Integrated Microbial Genomes (IMG) system is a comprehensive data management system that supports multidimensional comparative analysis of genomic data. At the core of the IMG system is a data warehouse that contains genome and metagenome datasets sequenced at the Joint Genome Institute or provided by scientific users, as well as public genome datasets available at the National Center for Biotechnology Information Genbank sequence data archive. Genomes and metagenome datasets are processed using IMG's microbial genome and metagenome sequence data processing pipelines and are integrated into the data warehouse using IMG's data integration toolkits. Microbial genome and metagenome application specific data marts and user interfaces provide access to different subsets of IMG's data and analysis toolkits. This review article revisits IMG's original aims, highlights key milestones reached by the system during the past 10 years, and discusses the main challenges faced by a rapidly expanding system, in particular the complexity of maintaining such a system in an academic setting with limited budgets and computing and data management infrastructure. Copyright © 2015 Elsevier Ltd. All rights reserved.

SATCHMO-JS: a webserver for simultaneous protein multiple sequence alignment and phylogenetic tree construction.

PubMed

Hagopian, Raffi; Davidson, John R; Datta, Ruchira S; Samad, Bushra; Jarvis, Glen R; Sjölander, Kimmen

2010-07-01

We present the jump-start simultaneous alignment and tree construction using hidden Markov models (SATCHMO-JS) web server for simultaneous estimation of protein multiple sequence alignments (MSAs) and phylogenetic trees. The server takes as input a set of sequences in FASTA format, and outputs a phylogenetic tree and MSA; these can be viewed online or downloaded from the website. SATCHMO-JS is an extension of the SATCHMO algorithm, and employs a divide-and-conquer strategy to jump-start SATCHMO at a higher point in the phylogenetic tree, reducing the computational complexity of the progressive all-versus-all HMM-HMM scoring and alignment. Results on a benchmark dataset of 983 structurally aligned pairs from the PREFAB benchmark dataset show that SATCHMO-JS provides a statistically significant improvement in alignment accuracy over MUSCLE, Multiple Alignment using Fast Fourier Transform (MAFFT), ClustalW and the original SATCHMO algorithm. The SATCHMO-JS webserver is available at http://phylogenomics.berkeley.edu/satchmo-js. The datasets used in these experiments are available for download at http://phylogenomics.berkeley.edu/satchmo-js/supplementary/.

Strong genetic structure corresponds to small-scale geographic breaks in the Australian alpine grasshopper Kosciuscola tristis.

PubMed

Slatyer, Rachel A; Nash, Michael A; Miller, Adam D; Endo, Yoshinori; Umbers, Kate D L; Hoffmann, Ary A

2014-10-02

Mountain landscapes are topographically complex, creating discontinuous 'islands' of alpine and sub-alpine habitat with a dynamic history. Changing climatic conditions drive their expansion and contraction, leaving signatures on the genetic structure of their flora and fauna. Australia's high country covers a small, highly fragmented area. Although the area is thought to have experienced periods of relative continuity during Pleistocene glacial periods, small-scale studies suggest deep lineage divergence across low-elevation gaps. Using both DNA sequence data and microsatellite markers, we tested the hypothesis that genetic partitioning reflects observable geographic structuring across Australia's mainland high country, in the widespread alpine grasshopper Kosciuscola tristis (Sjösted). We found broadly congruent patterns of regional structure between the DNA sequence and microsatellite datasets, corresponding to strong divergence among isolated mountain regions. Small and isolated mountains in the south of the range were particularly distinct, with well-supported divergence corresponding to climate cycles during the late Pliocene and Pleistocene. We found mixed support, however, for divergence among other mountain regions. Interestingly, within areas of largely contiguous alpine and sub-alpine habitat around Mt Kosciuszko, microsatellite data suggested significant population structure, accompanied by a strong signature of isolation-by-distance. Consistent patterns of strong lineage divergence among different molecular datasets indicate genetic breaks between populations inhabiting geographically distinct mountain regions. Three primary phylogeographic groups were evident in the highly fragmented Victorian high country, while within-region structure detected with microsatellites may reflect more recent population isolation. Despite the small area of Australia's alpine and sub-alpine habitats, their low topographic relief and lack of extensive glaciation, divergence among populations was on the same scale as that detected in much more extensive Northern hemisphere mountain systems. The processes driving divergence in the Australian mountains might therefore differ from their Northern hemisphere counterparts.

Ultra-deep mutant spectrum profiling: improving sequencing accuracy using overlapping read pairs.

PubMed

Chen-Harris, Haiyin; Borucki, Monica K; Torres, Clinton; Slezak, Tom R; Allen, Jonathan E

2013-02-12

High throughput sequencing is beginning to make a transformative impact in the area of viral evolution. Deep sequencing has the potential to reveal the mutant spectrum within a viral sample at high resolution, thus enabling the close examination of viral mutational dynamics both within- and between-hosts. The challenge however, is to accurately model the errors in the sequencing data and differentiate real viral mutations, particularly those that exist at low frequencies, from sequencing errors. We demonstrate that overlapping read pairs (ORP) -- generated by combining short fragment sequencing libraries and longer sequencing reads -- significantly reduce sequencing error rates and improve rare variant detection accuracy. Using this sequencing protocol and an error model optimized for variant detection, we are able to capture a large number of genetic mutations present within a viral population at ultra-low frequency levels (<0.05%). Our rare variant detection strategies have important implications beyond viral evolution and can be applied to any basic and clinical research area that requires the identification of rare mutations.

Gait Recognition Based on Convolutional Neural Networks

NASA Astrophysics Data System (ADS)

Sokolova, A.; Konushin, A.

2017-05-01

In this work we investigate the problem of people recognition by their gait. For this task, we implement deep learning approach using the optical flow as the main source of motion information and combine neural feature extraction with the additional embedding of descriptors for representation improvement. In order to find the best heuristics, we compare several deep neural network architectures, learning and classification strategies. The experiments were made on two popular datasets for gait recognition, so we investigate their advantages and disadvantages and the transferability of considered methods.