Abstract:
BACKGROUND:Visualization tools for deep learning models typically focus on discovering key input features without considering how such low level features are combined in intermediate layers to make decisions. Moreover, many of these methods examine a network's response to specific input examples that may be insufficient to reveal the complexity of model decision making. RESULTS:We present DeepResolve, an analysis framework for deep convolutional models of genome function that visualizes how input features contribute individually and combinatorially to network decisions. Unlike other methods, DeepResolve does not depend upon the analysis of a predefined set of inputs. Rather, it uses gradient ascent to stochastically explore intermediate feature maps to 1) discover important features, 2) visualize their contribution and interaction patterns, and 3) analyze feature sharing across tasks that suggests shared biological mechanism. We demonstrate the visualization of decision making using our proposed method on deep neural networks trained on both experimental and synthetic data. DeepResolve is competitive with existing visualization tools in discovering key sequence features, and identifies certain negative features and non-additive feature interactions that are not easily observed with existing tools. It also recovers similarities between poorly correlated classes which are not observed by traditional methods. DeepResolve reveals that DeepSEA's learned decision structure is shared across genome annotations including histone marks, DNase hypersensitivity, and transcription factor binding. We identify groups of TFs that suggest known shared biological mechanism, and recover correlation between DNA hypersensitivities and TF/Chromatin marks. CONCLUSIONS:DeepResolve is capable of visualizing complex feature contribution patterns and feature interactions that contribute to decision making in genomic deep convolutional networks. It also recovers feature sharing and class similarities which suggest interesting biological mechanisms. DeepResolve is compatible with existing visualization tools and provides complementary insights.
journal_name
BMC Bioinformaticsjournal_title
BMC bioinformaticsauthors
Liu G,Zeng H,Gifford DKdoi
10.1186/s12859-019-2957-4subject
Has Abstractpub_date
2019-07-19 00:00:00pages
401issue
1issn
1471-2105pii
10.1186/s12859-019-2957-4journal_volume
20pub_type
杂志文章abstract:BACKGROUND:The wide use of high-throughput DNA microarray technology provide an increasingly detailed view of human transcriptome from hundreds to thousands of genes. Although biomedical researchers typically design microarray experiments to explore specific biological contexts, the relationships between genes are hard...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-12-S5-S7
更新日期:2011-01-01 00:00:00
abstract:BACKGROUND:Analyzing Variance heterogeneity in genome wide association studies (vGWAS) is an emerging approach for detecting genetic loci involved in gene-gene and gene-environment interactions. vGWAS analysis detects variability in phenotype values across genotypes, as opposed to typical GWAS analysis, which detects v...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/s12859-018-2061-1
更新日期:2018-03-21 00:00:00
abstract:BACKGROUND:Community structure is ubiquitous in biological networks. There has been an increased interest in unraveling the community structure of biological systems as it may provide important insights into a system's functional components and the impact of local structures on dynamics at a global scale. Choosing an a...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-15-220
更新日期:2014-06-25 00:00:00
abstract:BACKGROUND:Metabolic networks reflect the relationships between metabolites (biomolecules) and the enzymes (proteins), and are of particular interest since they describe all chemical reactions of an organism. The metabolic networks are constructed from the genome sequence of an organism, and the graphs can be used to s...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/s12859-019-3112-y
更新日期:2019-10-15 00:00:00
abstract:BACKGROUND:The main research topic in this paper is how to compare multiple biological experiments using transcriptome data, where each experiment is measured and designed to compare control and treated samples. Comparison of multiple biological experiments is usually performed in terms of the number of DEGs in an arbi...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/s12859-019-3302-7
更新日期:2019-12-27 00:00:00
abstract:BACKGROUND:Gene expression patterns of olfactory receptors (ORs) are an important component of the signal encoding mechanism in the olfactory system since they determine the interactions between odorant ligands and sensory neurons. We have developed the Olfactory Receptor Microarray Database (ORMD) to house OR gene exp...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-8-231
更新日期:2007-06-30 00:00:00
abstract:BACKGROUND:We have previously described an approach to predicting the substrate specificity of serine-threonine protein kinases. The method, named Predikin, identifies key conserved substrate-determining residues in the kinase catalytic domain that contact the substrate in the region of the phosphorylation site and so ...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-9-245
更新日期:2008-05-26 00:00:00
abstract:BACKGROUND:In recent years, protein-protein interaction (PPI) networks have been well recognized as important resources to elucidate various biological processes and cellular mechanisms. In this paper, we address the problem of predicting protein complexes from a PPI network. This problem has two difficulties. One is r...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/s12859-017-1920-5
更新日期:2017-12-06 00:00:00
abstract:BACKGROUND:Designing small-molecule kinase inhibitors with desirable selectivity profiles is a major challenge in drug discovery. A high-throughput screen for inhibitors of a given kinase will typically yield many compounds that inhibit more than one kinase. A series of chemical modifications are usually required befor...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-9-491
更新日期:2008-11-25 00:00:00
abstract:BACKGROUND:As part of its broad and ambitious mission, the MicroArray Quality Control (MAQC) project reported the results of experiments using External RNA Controls (ERCs) on five microarray platforms. For most platforms, several different methods of data processing were considered. However, there was no similar consid...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-8-371
更新日期:2007-10-03 00:00:00
abstract:BACKGROUND:In real-time PCR data analysis, the cycle threshold (CT) method is currently the gold standard. This method is based on an assumption of equal PCR efficiency in all reactions, and precision may suffer if this condition is not met. Nonlinear regression analysis (NLR) or curve fitting has therefore been sugges...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-7-107
更新日期:2006-03-03 00:00:00
abstract:BACKGROUND:Protein-protein docking, which aims to predict the structure of a protein-protein complex from its unbound components, remains an unresolved challenge in structural bioinformatics. An important step is the ranking of docked poses using a scoring function, for which many methods have been developed. There is ...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-14-286
更新日期:2013-10-01 00:00:00
abstract:BACKGROUND:Aptamers are nucleic acids selected for their ability to bind to molecules of interest and may provide the basis for a whole new class of medicines. If the aptamer is simply a dsDNA molecule with a ssDNA overhang (a "sticky" end) then the segment of ssDNA that complements that overhang provides a known bindi...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-8-S7-S11
更新日期:2007-11-01 00:00:00
abstract:BACKGROUND:The adaptation of the CRISPR-Cas9 system to pooled library gene knockout screens in mammalian cells represents a major technological leap over RNA interference, the prior state of the art. New methods for analyzing the data and evaluating results are needed. RESULTS:We offer BAGEL (Bayesian Analysis of Gene...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/s12859-016-1015-8
更新日期:2016-04-16 00:00:00
abstract:BACKGROUND:In many bacteria, intragenomic diversity in synonymous codon usage among genes has been reported. However, no quantitative attempt has been made to compare the diversity levels among different genomes. Here, we introduce a mean dissimilarity-based index (Dmean) for quantifying the level of diversity in synon...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-10-167
更新日期:2009-06-01 00:00:00
abstract:BACKGROUND:Analyzing the amino acid sequence of an intrinsically disordered protein (IDP) in an evolutionary context can yield novel insights on the functional role of disordered regions and sequence element(s). However, in the case of many IDPs, the lack of evolutionary conservation of the primary sequence can hamper ...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/s12859-015-0592-2
更新日期:2015-05-13 00:00:00
abstract:BACKGROUND:The goal of class prediction studies is to develop rules to accurately predict the class membership of new samples. The rules are derived using the values of the variables available for each subject: the main characteristic of high-dimensional data is that the number of variables greatly exceeds the number o...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-11-523
更新日期:2010-10-20 00:00:00
abstract:BACKGROUND:The binding of regulatory proteins to their specific DNA targets determines the accurate expression of the neighboring genes. The in silico prediction of new binding sites in completely sequenced genomes is a key aspect in the deeper understanding of gene regulatory networks. Several algorithms have been des...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-11-554
更新日期:2010-11-11 00:00:00
abstract:BACKGROUND:Detection of genomic DNA copy number variations (CNVs) can provide a complete and more comprehensive view of human disease. It is interesting to identify and represent relevant CNVs from a genome-wide data due to high data volume and the complexity of interactions. RESULTS:In this paper, we incorporate the ...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-12-S5-S4
更新日期:2011-01-01 00:00:00
abstract:BACKGROUND:High-throughput technology allows for genome-wide measurements at different molecular levels for the same patient, e.g. single nucleotide polymorphisms (SNPs) and gene expression. Correspondingly, it might be beneficial to also integrate complementary information from different molecular levels when building...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/s12859-016-1183-6
更新日期:2016-08-30 00:00:00
abstract:BACKGROUND:Recent technological advances in DNA sequencing and genotyping have led to the accumulation of a remarkable quantity of data on genetic polymorphisms. However, the development of new statistical and computational tools for effective processing of these data has not been equally as fast. In particular, Machin...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-10-S6-S24
更新日期:2009-06-16 00:00:00
abstract:BACKGROUND:Current malaria diagnosis relies primarily on microscopic examination of Giemsa-stained thick and thin blood films. This method requires vigorously trained technicians to efficiently detect and classify the malaria parasite species such as Plasmodium falciparum (Pf) and Plasmodium vivax (Pv) for an appropria...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-13-S17-S18
更新日期:2012-01-01 00:00:00
abstract:BACKGROUND:The development of next-generation sequencing has made it possible to sequence whole genomes at a relatively low cost. However, de novo genome assemblies remain challenging due to short read length, missing data, repetitive regions, polymorphisms and sequencing errors. As more and more genomes are sequenced,...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/s12859-017-1911-6
更新日期:2017-11-10 00:00:00
abstract:BACKGROUND:Normalization is the process of removing non-biological sources of variation between array experiments. Recent investigations of data in gene expression databases for varying organisms and tissues have shown that the majority of expressed genes exhibit a power-law distribution with an exponent close to -1 (i...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-6-37
更新日期:2005-02-23 00:00:00
abstract:BACKGROUND:Sequence mutations represent a driving force of adaptive evolution in bacterial pathogens. It is especially evident in reductive genome evolution where bacteria underwent lifestyles shifting from a free-living to a strictly intracellular or host-depending life. It resulted in loss-of-function mutations and/o...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-10-S1-S3
更新日期:2009-01-30 00:00:00
abstract:BACKGROUND:Profile-based comparison of multiple sequence alignments is a powerful methodology for the detection remote protein sequence similarity, which is essential for the inference and analysis of protein structure, function, and evolution. Accurate estimation of statistical significance of detected profile similar...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-10-399
更新日期:2009-12-04 00:00:00
abstract:BACKGROUND:The imputation of missing values is necessary for the efficient use of DNA microarray data, because many clustering algorithms and some statistical analysis require a complete data set. A few imputation methods for DNA microarray data have been introduced, but the efficiency of the methods was low and the va...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-5-160
更新日期:2004-10-26 00:00:00
abstract:BACKGROUND:The information in large collections of phylogenetic trees is useful for many comparative genomic studies. Therefore, there is a need for flexible tools that allow exploration of such collections in order to retrieve relevant data as quickly as possible. RESULTS:In this paper, we present TPMS (Tree Pattern-...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-14-109
更新日期:2013-03-27 00:00:00
abstract:BACKGROUND:Antibodies revolutionized cancer treatment over the past decades. Despite their successfully application, there are still challenges to overcome to improve efficacy, such as the heterogeneous distribution of antibodies within tumors. Tumor microenvironment features, such as the distribution of tumor and othe...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/s12859-020-03854-2
更新日期:2020-11-17 00:00:00
abstract:BACKGROUND:Many proteins contain conserved sequence patterns (motifs) that contribute to their functionality. The process of experimentally identifying and validating novel protein motifs can be difficult, expensive, and time consuming. A means for helping to identify in advance the possible function of a novel motif i...
journal_title:BMC bioinformatics
pub_type: 杂志文章
doi:10.1186/1471-2105-12-379
更新日期:2011-09-26 00:00:00