Search

We present a new combinatorial model for identifying regulatory modules in gene co-expression data using a decomposition into weighted cliques. To capture complex interaction effects, we generalize the previously-studied weighted edge clique partition problem. As a first step, we restrict ourselves to the noise-free setting, and show that the problem is fixed parameter tractable when parameterized by the number of modules (cliques). We present two new algorithms for finding these decompositions, using linear programming and integer partitioning to determine the clique weights. Further, we implement these algorithms in Python and test them on a biologically-inspired synthetic corpus generated using real-world data from transcription factors and a latent variable analysis of co-expression in varying cell types.

To make machine learning analyses in the life sciences more computationally reproducible, we propose standards based on data, model, and code publication, programming best practices, and workflow automation. By meeting these standards, the community of researchers applying machine learning methods in the life sciences can ensure that their analyses are worthy of trust.

Background Pooling cells from multiple biological samples prior to library preparation within the same single-cell RNA sequencing experiment provides several advantages, including lower library preparation costs and reduced unwanted technological variation, such as batch effects. Computational demultiplexing tools based on natural genetic variation between individuals provide a simple approach to demultiplex samples, which does not require complex additional experimental procedures. However, to our knowledge these tools have not been evaluated in cancer, where somatic variants, which could differ between cells from the same sample, may obscure the signal in natural genetic variation. Results Here, we performed in silico benchmark evaluations by combining raw sequencing reads from multiple single-cell samples in high-grade serous ovarian cancer, which has a high copy number burden, and lung adenocarcinoma, which has a high tumor mutational burden. Our results confirm that genetic demultiplexing tools can be effectively deployed on cancer tissue using a pooled experimental design, although high proportions of ambient RNA from cell debris reduce performance. Conclusions This strategy provides significant cost savings through pooled library preparation. To facilitate similar analyses at the experimental design phase, we provide freely accessible code and a reproducible Snakemake workflow built around the best-performing tools found in our in silico benchmark evaluations, available at https://github.com/lmweber/snp-dmx-cancer.

Most drugs entering clinical trials fail, often related to an incomplete understanding of the mechanisms governing drug response. Machine learning techniques hold immense promise for better drug response predictions, but most have not reached clinical practice due to their lack of interpretability and their focus on monotherapies. We address these challenges by developing DrugCell, an interpretable deep learning model of human cancer cells trained on the responses of 1,235 tumor cell lines to 684 drugs. Tumor genotypes induce states in cellular subsystems that are integrated with drug structure to predict response to therapy and, simultaneously, learn biological mechanisms underlying the drug response. DrugCell predictions are accurate in cell lines and also stratify clinical outcomes. Analysis of DrugCell mechanisms leads directly to the design of synergistic drug combinations, which we validate systematically by combinatorial CRISPR, drug-drug screening in vitro, and patient-derived xenografts. DrugCell provides a blueprint for constructing interpretable models for predictive medicine. Graphical Abstract

Genomic data sharing accelerates research. Data are most valuable when they are accompanied by detailed metadata. To date, metadata are often human-annotated descriptions of samples and their handling. We discuss how machine learning-derived elements complement such descriptions to enhance the research ecosystem around genomic data.

Knowledge graphs can support many biomedical applications. These graphs represent biomedical concepts and relationships in the form of nodes and edges. In this review, we discuss how these graphs are constructed and applied with a particular focus on how machine learning approaches are changing these processes. Biomedical knowledge graphs have often been constructed by integrating databases that were populated by experts via manual curation, but we are now seeing a more robust use of automated systems. A number of techniques are used to represent knowledge graphs, but often machine learning methods are used to construct a low-dimensional representation that can support many different applications. This representation is designed to preserve a knowledge graph's local and/or global structure. Additional machine learning methods can be applied to this representation to make predictions within genomic, pharmaceutical, and clinical domains. We frame our discussion first around knowledge graph construction and then around unifying representational learning techniques and unifying applications. Advances in machine learning for biomedicine are creating new opportunities across many domains, and we note potential avenues for future work with knowledge graphs that appear particularly promising.

Genetic interactions have been recognized as a potentially important contributor to the heritability of complex diseases. Nevertheless, due to small effect sizes and stringent multiple-testing correction, identifying genetic interactions in complex diseases is particularly challenging. To address the above challenges, many genomic research initiatives collaborate to form large-scale consortia and develop open access to enable sharing of genome-wide association study (GWAS) data. Despite the perceived benefits of data sharing from large consortia, a number of practical issues have arisen, such as privacy concerns on individual genomic information and heterogeneous data sources from distributed GWAS databases. In the context of large consortia, we demonstrate that the heterogeneously appearing marginal effects over distributed GWAS databases can offer new insights into genetic interactions for which conventional methods have had limited success. In this paper, we develop a novel two-stage testing procedure, named phylogenY-based Effect-size Tests for Interactions using first 2 moments (YETI2), to detect genetic interactions through both pooled marginal effects, in terms of averaging site-specific marginal effects, and heterogeneity in marginal effects across sites, using a meta-analytic framework. YETI2 can not only be applied to large consortia without shared personal information, but also can be used to leverage underlying heterogeneity in marginal effects to prioritize potential genetic interactions. We investigate the performance of YETI2 through simulation studies, and apply YETI2 to a bladder cancer data from dbGaP.

Pathway and cell type signatures are patterns present in transcriptome data that are associated with biological processes or phenotypic consequences. These signatures result from specific cell type and pathway expression but can require large transcriptomic compendia to detect. Machine learning techniques can be powerful tools for signature discovery through their ability to provide accurate and interpretable results. In this review, we discuss various machine learning applications to extract pathway and cell type signatures from transcriptomic compendia. We focus on the biological motivations and interpretation for both supervised and unsupervised learning approaches in this setting. We consider recent advances, including deep learning, and their applications to expanding bulk and single-cell RNA data. As data and computational resources increase, there will be more opportunities for machine learning to aid in revealing biological signatures.

Single-cell RNA-sequencing (scRNA-seq) has made it possible to profile gene expression in tissues at high resolution. An important preprocessing step prior to performing downstream analyses is to identify and remove cells with poor or degraded sample quality using quality control (QC) metrics. Two widely used QC metrics to identify a ‘low-quality’ cell are (i) if the cell includes a high proportion of reads that map to mitochondrial DNA (mtDNA) encoded genes and (ii) if a small number of genes are detected. Current best practices use these QC metrics independently with either arbitrary, uniform thresholds (e.g. 5%) or biological context-dependent (e.g. species) thresholds, and fail to jointly model these metrics in a data-driven manner. Current practices are often overly stringent and especially untenable on certain types of tissues, such as archived tumor tissues, or tissues associated with mitochondrial function, such as kidney tissue [1]. We propose a data-driven QC metric (miQC) that jointly models both the proportion of reads mapping to mtDNA genes and the number of detected genes with mixture models in a probabilistic framework to predict the low-quality cells in a given dataset. We demonstrate how our QC metric easily adapts to different types of single-cell datasets to remove low-quality cells while preserving high-quality cells that can be used for downstream analyses. Our software package is available at https://bioconductor.org/packages/miQC., Author summary We developed the miQC package to predict the low-quality cells in a given scRNA-seq dataset by jointly modeling both the proportion of reads mapping to mitochondrial DNA (mtDNA) genes and the number of detected genes using mixture models in a probabilistic framework. We demonstrate how our QC metric easily adapts to different types of single-cell datasets to remove low-quality cells while preserving high-quality cells that can be used for downstream analyses.

Preprints allow researchers to make their findings available to the scientific community before they have undergone peer review. Studies on preprints within bioRxiv have been largely focused on article metadata and how often these preprints are downloaded, cited, published, and discussed online. A missing element that has yet to be examined is the language contained within the bioRxiv preprint repository. We sought to compare and contrast linguistic features within bioRxiv preprints to published biomedical text as a whole as this is an excellent opportunity to examine how peer review changes these documents. The most prevalent features that changed appear to be associated with typesetting and mentions of supplementary sections or additional files. In addition to text comparison, we created document embeddings derived from a preprint-trained word2vec model. We found that these embeddings are able to parse out different scientific approaches and concepts, link unannotated preprint-peer reviewed article pairs, and identify journals that publish linguistically similar papers to a given preprint. We also used these embeddings to examine factors associated with the time elapsed between the posting of a first preprint and the appearance of a peer reviewed publication. We found that preprints with more versions posted and more textual changes took longer to publish. Lastly, we constructed a web application (https://greenelab.github.io/preprint-similarity-search/) that allows users to identify which journals and articles that are most linguistically similar to a bioRxiv or medRxiv preprint as well as observe where the preprint would be positioned within a published article landscape.

Computational methods have reshaped the landscape of modern biology. While the biomedical community is increasingly dependent on computational tools, the mechanisms ensuring open data, open software, and reproducibility are variably enforced by academic institutions, funders, and publishers. Publications may present academic software for which essential materials are or become unavailable, such as source code and documentation. Publications that lack such information compromise the role of peer review in evaluating technical strength and scientific contribution. Incomplete ancillary information for an academic software package may bias or limit any subsequent work produced with the tool. We provide eight recommendations across four different domains to improve reproducibility, transparency, and rigor in computational biology - precisely on the main values which should be emphasized in life science curricula. Our recommendations for improving software availability, usability, and archival stability aim to foster a sustainable data science ecosystem in biomedicine and life science research.

Single-cell RNA sequencing (scRNA-seq) is a powerful tool to simultaneously sequencing the transcriptomes of a large number of individual cells at a high resolution. These data usually contain measurements of gene expression for many genes in thousands or tens of thousands of cells, though some datasets now reach the million-cell mark. Projecting high dimensional scRNA-seq data into a low dimensional space is essential for downstream analysis and data visualization. Many recent preprints accomplish this using variational autoencoders (VAE), generative models that learn underlying structure of data by compress it into a constrained, low dimensional space. The low dimensional spaces generated by VAEs have revealed complex patterns and novel biological signals from large-scale gene expression data and drug response predictions. Here, we evaluate a simple VAE approach for gene expression data, Tybalt, by training and measuring its performance on sets of simulated scRNA-seq data. We find a number of counter-intuitive performance changes: i.e., deeper neural networks can struggle when datasets contain more observations under some parameter configurations. We show that these methods are highly sensitive to parameter tuning: when tuned, the performance of the Tybalt model, which was not optimized for scRNA-seq data, outperforms other popular dimension reduction approaches - PCA, ZIFA, UMAP and t-SNE. On the other hand, without tuning performance can also be remarkably poor on the same data. Our results should discourage authors and reviewers from relying on empirical performance comparisons to evaluate the relative value of contributions in this area at this time. Instead, we recommend that attempts to compare or benchmark autoencoder methods for scRNA-seq data be performed by disinterested third parties or by methods developers only on unseen benchmark data that are provided to all participants simultaneously because the potential for performance differences due to unequal parameter tuning is so high.

The Kipoi repository provides a home for trained machine-learning models that facilitates their sharing and reuse.

Replication, validation and extension of experiments are crucial for scientific progress. Computational experiments are inherently scriptable and should be easy to reproduce. However, it remains difficult and time consuming to reproduce computational results because analyses are designed and run in a specific computing environment, which may be difficult or impossible to match from written instructions. We report a workflow named continuous analysis that can build reproducibility into computational analyses. Continuous analysis combines Docker, a container technology akin to virtual machines, with continuous integration, a software development technique, to automatically re-run a computational analysis whenever updates or improvements are made to source code or data. This allows results to be accurately reproduced without needing to contact the study authors. Continuous analysis allows reviewers, editors or readers to verify reproducibility without manually downloading and re-running any code and can provide an audit trail for analyses of data that cannot be shared.

Recent work suggests that gene expression dependencies can be predicted almost as well by using random networks as by using experimentally derived interaction networks. We hypothesize that this effect is highly variable across genes, as useful and robust experimental evidence exists for some genes but not others. To explore this variation, we take the k-core decomposition of the STRING network, and compare it to a degree-matched random model. We show that when low-degree nodes are removed, expression dependencies in the remaining genes can be predicted better by the resulting network than by the random model.

In biomedical applications of machine learning, relevant information often has a rich structure that is not easily encoded as real-valued predictors. Examples of such data include DNA or RNA sequences, gene sets or pathways, gene interaction or coexpression networks, ontologies, and phylogenetic trees. We highlight recent examples of machine learning models that use structure to constrain model architecture or incorporate structured data into model training. For machine learning in biomedicine, where sample size is limited and model interpretability is critical, incorporating prior knowledge in the form of structured data can be particularly useful. The area of research would benefit from performant open source implementations and independent benchmarking efforts., Comments welcome at https://greenelab.github.io/biopriors-review/

Background: Data sharing accelerates scientific progress but sharing individual-level data while preserving patient privacy presents a barrier. Methods and Results: Using pairs of deep neural networks, we generated simulated, synthetic participants that closely resemble participants of the SPRINT trial (Systolic Blood Pressure Trial). We showed that such paired networks can be trained with differential privacy, a formal privacy framework that limits the likelihood that queries of the synthetic participants’ data could identify a real a participant in the trial. Machine learning predictors built on the synthetic population generalize to the original data set. This finding suggests that the synthetic data can be shared with others, enabling them to perform hypothesis-generating analyses as though they had the original trial data. Conclusions: Deep neural networks that generate synthetic participants facilitate secondary analyses and reproducible investigation of clinical data sets by enhancing data sharing while preserving participant privacy.

The Critical Assessment of Functional Annotation (CAFA) is an ongoing, global, community-driven effort to evaluate and improve the computational annotation of protein function. Here we report on the results of the third CAFA challenge, CAFA3, that featured an expanded analysis over the previous CAFA rounds, both in terms of volume of data analyzed and the types of analysis performed. In a novel and major new development, computational predictions and assessment goals drove some of the experimental assays, resulting in new functional annotations for more than 1000 genes. Specifically, we performed experimental whole-genome mutation screening in Candida albicans and Pseudomonas aureginosa genomes, which provided us with genome-wide experimental data for genes associated with biofilm formation and motility (P. aureginosa only). We further performed targeted assays on selected genes in Drosophila melanogaster, which we suspected of being involved in long-term memory. We conclude that, while predictions of the molecular function and biological process annotations have slightly improved over time, those of the cellular component have not. Term-centric prediction of experimental annotations remains equally challenging; although the performance of the top methods is significantly better than expectations set by baseline methods in C. albicans and D. melanogaster, it leaves considerable room and need for improvement. We finally report that the CAFA community now involves a broad range of participants with expertise in bioinformatics, biological experimentation, biocuration, and bioontologies, working together to improve functional annotation, computational function prediction, and our ability to manage big data in the era of large experimental screens.

BackgroundUnsupervised compression algorithms applied to gene expression data extract latent, or hidden, signals representing technical and biological sources of variation. However, these algorithms require a user to select a biologically-appropriate latent dimensionality. In practice, most researchers select a single algorithm and latent dimensionality. We sought to determine the extent by which using multiple dimensionalities across ensemble compression models improves biological representations.ResultsWe compressed gene expression data from three large datasets consisting of adult normal tissue, adult cancer tissue, and pediatric cancer tissue. We compressed these data into many latent dimensionalities ranging from 2 to 200. We observed various tradeoffs across latent dimensionalities and compression models. For example, we observed high model stability between principal components analysis (PCA), independent components analysis (ICA), and non-negative matrix factorization (NMF). We identified more unique biological signatures in ensembles of denoising autoencoder (DAE) and variational autoencoder (VAE) models in intermediate latent dimensionalities. However, we captured the most pathway-associated features using all compressed features across algorithms and dimensionalities. Optimized at different latent dimensionalities, compression models detect generalizable gene expression signatures representing sex, neuroblastoma MYCN amplification, and cell types. In two supervised machine learning tasks, compressed features optimized predictions at different latent dimensionalities.ConclusionsThere is no single best latent dimensionality or compression algorithm for analyzing gene expression data. Instead, using feature ensembles from different compression models across latent space dimensionalities optimizes biological representations.

Integrative networks combine multiple layers of biological data into a model of how genes work together to carry out cellular processes. Such networks become more valuable as they become more context-specific, for example, by capturing how genes work together in a certain tissue or cell type. We discuss the applications of these networks to the study of human disease. Once constructed, these networks provide the means to identify broad biological patterns underlying genes associated with complex traits and diseases. We cover the different types of integrative networks that currently exist, and how such networks that encompass multiple biological layers are constructed. We highlight how specificity can be incorporated into the reconstruction of different types of biomolecular interactions between genes, using tissue specificity as a motivating example. We discuss examples of cases where networks have been applied to study human diseases and opportunities for new applications. Integrative networks with specificity to tissue or other biological features provide new capabilities to researchers engaged in the study of human disease. We expect improved data and algorithms to continue and improve such networks, allowing them to provide more detailed and mechanistic predictions into the context-specific genetic etiology of common diseases.

A major bottleneck to our understanding of the genetic and molecular foundation of life lies in the ability to assign function to a gene and, subsequently, a protein. Traditional molecular and genetic experiments can provide the most reliable forms of identification, but are generally low-throughput, making such discovery and assignment a daunting task. The bottleneck has led to an increasing role for computational approaches. The Critical Assessment of Functional Annotation (CAFA) effort seeks to measure the performance of computational methods. In CAFA3 we performed selected screens, including an effort focused on long-term memory. We used homology and previous CAFA predictions to identify 29 keyDrosophilagenes, which we tested via a long-term memory screen. We identify 11 novel genes that are involved in long-term memory formation and show a high level of connectivity with previously identified learning and memory genes. Our study provides first higher-order behavioral assay and organism screen used for CAFA assessments and revealed previously uncharacterized roles of multiple genes as possible regulators of neuronal plasticity at the boundary of information acquisition and memory formation.

SUMMARYUnsupervised machine learning methods provide a promising means to analyze and interpret large datasets. However, most gene expression datasets generated by individual researchers remain too small to fully benefit from these methods. In the case of rare diseases, there may be too few cases available, even when multiple studies are combined. We trained a Pathway Level Information ExtractoR (PLIER) model using on a large public data compendium comprised of multiple experiments, tissues, and biological conditions. We then transferred the model to small rare disease datasets in an approach we term MultiPLIER. Models constructed from large, diverse public data i) included features that aligned well to important biological factors; ii) were more comprehensive than those constructed from individual datasets or conditions; iii) transferred to rare disease datasets where the models describe biological processes related to disease severity more effectively than models trained on specifically those datasets.

BackgroundInvestigators often interpret genome-wide data by analyzing the expression levels of genes within pathways. While this within-pathway analysis is routine, the products of any one pathway can affect the activity of other pathways. Past efforts to identify relationships between biological processes have evaluated overlap in knowledge bases or evaluated changes that occur after specific treatments. Individual experiments can highlight condition-specific pathway-pathway relationships; however, constructing a complete network of such relationships across many conditions requires analyzing results from many studies.ResultsWe developed PathCORE-T framework by implementing existing methods to identify pathway-pathway transcriptional relationships evident across a broad data compendium. PathCORE-T is applied to the output of feature construction algorithms; it identifies pairs of pathways observed in features more than expected by chance as functionally co-occurring. We demonstrate PathCORE-T by analyzing an existing eADAGE model of a microbial compendium and building and analyzing NMF features from the TCGA dataset of 33 cancer types. The PathCORE-T framework includes a demonstration web interface, with source code, that users can launch to (1) visualize the network and (2) review the expression levels of associated genes in the original data. PathCORE-T creates and displays the network of globally co-occurring pathways based on features observed in a machine learning analysis of gene expression data.ConclusionsThe PathCORE-T framework identifies transcriptionally co-occurring pathways from the results of unsupervised analysis of gene expression data and visualizes the relationships between pathways as a network. PathCORE-T recapitulated previously described pathway-pathway relationships and suggested experimentally testable additional hypotheses that remain to be explored.

Motivation Decreasing costs are making it feasible to perform time series proteomics and genomics experiments with more replicates and higher resolution than ever before. With more replicates and time points, proteome and genome-wide patterns of expression are more readily discernible. These larger experiments require more batches exacerbating batch effects and increasing the number of bias trends. In the case of proteomics, where methods frequently result in missing data this increasing scale is also decreasing the number of peptides observed in all samples. The sources of batch effects and missing data are incompletely understood necessitating novel techniques. Results Here we show that by exploiting the structure of time series experiments, it is possible to accurately and reproducibly model and remove batch effects. We implement Learning and Imputation for Mass-spec Bias Reduction (LIMBR) software, which builds on previous block-based models of batch effects and includes features specific to time series and circadian studies. To aid in the analysis of time series proteomics experiments, which are often plagued with missing data points, we also integrate an imputation system. By building LIMBR for imputation and time series tailored bias modeling into one straightforward software package, we expect that the quality and ease of large-scale proteomics and genomics time series experiments will be significantly increased. Availability and implementation Python code and documentation is available for download at https://github.com/aleccrowell/LIMBR and LIMBR can be downloaded and installed with dependencies using ‘pip install limbr’. Supplementary information Supplementary data are available at Bioinformatics online.

Precision medicine will revolutionize the way we treat and prevent disease. A major barrier to the implementation of precision medicine that clinicians and translational scientists face is understanding the underlying mechanisms of disease. We are starting to address this challenge through automatic approaches for information extraction, representation and analysis. Recent advances in text and data mining have been applied to a broad spectrum of key biomedical questions in genomics, pharmacogenomics and other fields. We present an overview of the fundamental methods for text and data mining, as well as recent advances and emerging applications toward precision medicine.

Modern technologies are capable of generating enormous amounts of data that measure complex biological systems. Computational biologists and bioinformatics scientists are increasingly being asked to use these data to reveal key systems-level properties. We review the extent to which curricula are changing in the era of big data. We identify key competencies that scientists dealing with big data are expected to possess across fields, and we use this information to propose courses to meet these growing needs. While bioinformatics programs have traditionally trained students in data-intensive science, we identify areas of particular biological, computational and statistical emphasis important for this era that can be incorporated into existing curricula. For each area, we propose a course structured around these topics, which can be adapted in whole or in parts into existing curricula. In summary, specific challenges associated with big data provide an important opportunity to update existing curricula, but we do not foresee a wholesale redesign of bioinformatics training programs.

Omics data contains signal from the molecular, physical, and kinetic inter- and intra-cellular interactions that control biological systems. Matrix factorization techniques can reveal low-dimensional structure from high-dimensional data that reflect these interactions. These techniques can uncover new biological knowledge from diverse high-throughput omics data in topics ranging from pathway discovery to time course analysis. We review exemplary applications of matrix factorization for systems-level analyses. We discuss appropriate application of these methods, their limitations, and focus on analysis of results to facilitate optimal biological interpretation. The inference of biologically relevant features with matrix factorization enables discovery from high-throughput data beyond the limits of current biological knowledge—answering questions from high-dimensional data that we have not yet thought to ask.

The Cancer Genome Atlas (TCGA) has profiled over 10,000 tumors across 33 different cancer-types for many genomic features, including gene expression levels. Gene expression measurements capture substantial information about the state of each tumor. Certain classes of deep neural network models are capable of learning a meaningful latent space. Such a latent space could be used to explore and generate hypothetical gene expression profiles under various types of molecular and genetic perturbation. For example, one might wish to use such a model to predict a tumor’s response to specific therapies or to characterize complex gene expression activations existing in differential proportions in different tumors. Variational autoencoders (VAEs) are a deep neural network approach capable of generating meaningful latent spaces for image and text data. In this work, we sought to determine the extent to which a VAE can be trained to model cancer gene expression, and whether or not such a VAE would capture biologically-relevant features. In the following report, we introduce a VAE trained on TCGA pan-cancer RNA-seq data, identify specific patterns in the VAE encoded features, and discuss potential merits of the approach. We name our method “Tybalt” after an instigative, cat-like character who sets a cascading chain of events in motion in Shakespeare’s “Romeo and Juliet”. From a systems biology perspective, Tybalt could one day aid in cancer stratification or predict specific activated expression patterns that would result from genetic changes or treatment effects.

Differential expression experiments or other analyses often end in a list of genes. Pathway enrichment analysis is one method to discern important biological signals and patterns from noisy expression data. However, pathway enrichment analysis may perform suboptimally in situations where there are multiple implicated pathways – such as in the case of genes that define subtypes of complex diseases. Our simulation study shows that in this setting, standard overrepresentation analysis identifies many false positive pathways along with the true positives. These false positives hamper investigators’ attempts to glean biological insights from enrichment analysis. We develop and evaluate an approach that combines community detection over functional networks with pathway enrichment to reduce false positives. Our simulation study demonstrates that a large reduction in false positives can be obtained with a small decrease in power. Though we hypothesized that multiple communities might underlie previously described subtypes of high-grade serous ovarian cancer and applied this approach, our results do not support this hypothesis. In summary, applying community detection before enrichment analysis may ease interpretation for complex gene sets that represent multiple distinct pathways.

Motivation: Large compendia of gene expression data have proven valuable for the discovery of novel biological relationships. The majority of available RNA assays are run on microarray, while RNA-seq is becoming the platform of choice for new experiments. The data structure and distributions between the platforms differ, making it challenging to combine them. We performed supervised and unsupervised machine learning evaluations, as well as differential expression analyses, to assess which normalization methods are best suited for combining microarray and RNA-seq data. Results: We find that quantile and Training Distribution Matching normalization allow for supervised and unsupervised model training on microarray and RNA-seq data simultaneously. Nonparanormal normalization and z-scores are also appropriate for some applications, including differential expression analysis. Availability and Implementation: These analyses were performed in R and are available at https://www.github.com/greenelab/RNAseq_titration_results under a BSD-3 clause license.

Recent technological advances allow for high throughput profiling of biological systems in a cost-efficient manner. The low cost of data generation is leading us to the "big data" era. The availability of big data provides unprecedented opportunities but also raises new challenges for data mining and analysis. In this review, we introduce key concepts in the analysis of big data, including both "machine learning" algorithms as well as "unsupervised" and "supervised" examples of each. We note packages for the R programming language that are available to perform machine learning analyses. In addition to programming based solutions, we review webservers that allow users with limited or no programming background to perform these analyses on large data compendia.

A recent approach for single-cell RNA-sequencing data uses Bayesian deep learning to correct technical artifacts and enable accurate and multifaceted downstream analyses.

Open, collaborative research is a powerful paradigm that can immensely strengthen the scientific process by integrating broad and diverse expertise. However, traditional research and multi-author writing processes break down at scale. We present new software named Manubot, available at https://manubot.org, to address the challenges of open scholarly writing. Manubot adopts the contribution workflow used by many large-scale open source software projects to enable collaborative authoring of scholarly manuscripts. With Manubot, manuscripts are written in Markdown and stored in a Git repository to precisely track changes over time. By hosting manuscript repositories publicly, such as on GitHub, multiple authors can simultaneously propose and review changes. A cloud service automatically evaluates proposed changes to catch errors. Publication with Manubot is continuous: When a manuscript’s source changes, the rendered outputs are rebuilt and republished to a web page. Manubot automates bibliographic tasks by implementing citation by identifier, where users cite persistent identifiers (e.g. DOIs, PubMed IDs, ISBNs, URLs), whose metadata is then retrieved and converted to a user-specified style. Manubot modernizes publishing to align with the ideals of open science by making it transparent, reproducible, immediate, versioned, collaborative, and free of charge., Author summary Traditionally, scholarly manuscripts have been written in private by a predefined team of collaborators. But now the internet enables realtime open science, where project communication occurs online in a public venue and anyone is able to contribute. Dispersed teams of online contributors require new tools to jointly prepare manuscripts. Existing tools fail to scale beyond tens of authors and struggle to support iterative refinement of proposed changes. Therefore, we created a system called Manubot for writing manuscripts based on collaborative version control. Manubot adopts the workflow from open source software development, which has enabled hundreds of contributors to simultaneously develop complex codebases such as Python and Linux, and applies it to open collaborative writing. Manubot also addresses other shortcomings of current publishing tools. Specifically, all changes to a manuscript are tracked, enabling transparency and better attribution of credit. Manubot automates many tasks, including creating the bibliography and deploying the manuscript as a webpage. Manubot webpages preserve old versions and provide a simple yet interactive interface for reading. As such, Manubot is a suitable foundation for next-generation preprints. Manuscript readers have ample opportunity to not only provide public peer review but also to contribute improvements, before and after journal publication.

Reproducing experiments is vital to science. Being able to replicate, validate and extend previous work also speeds new research projects. Reproducing computational biology experiments, which are scripted, should be straightforward. But reproducing such work remains challenging and time consuming. In the ideal world we would be able to quickly and easily rewind to the precise computing environment where results were generated. We would then be able to reproduce the original analysis or perform new analyses. We introduce a process termed “continuous analysis” which provides inherent reproducibility to computational research at a minimal cost to the researcher. Continuous analysis combines Docker, a container service similar to virtual machines, with continuous integration, a popular software development technique, to automatically re-run computational analysis whenever relevant changes are made to the source code. This allows results to be reproduced quickly, accurately and without needing to contact the original authors. Continuous analysis also provides an audit trail for analyses that use data with sharing restrictions. This allows reviewers, editors, and readers to verify reproducibility without manually downloading and rerunning any code. Example configurations are available at our online repository (https://github.com/greenelab/continuous_analysis).

BackgroundThe adoption of new bioinformatics webservers provides biological researchers with new analytical opportunities but also raises workflow challenges. These challenges include sharing collections of genes with collaborators, translating gene identifiers to the most appropriate nomenclature for each server, tracking these collections across multiple analysis tools and webservers, and maintaining effective records of the genes used in each analysis.DescriptionIn this paper, we present the Tribe webserver (available at https://tribe.greenelab.com), which addresses these challenges in order to make multi-server workflows seamless and reproducible. This allows users to create analysis pipelines that use their own sets of genes in combinations of specialized data mining webservers and tools while seamlessly maintaining gene set version control. Tribe’s web interface facilitates collaborative editing: users can share with collaborators, who can then view, download, and edit these collections. Tribe’s fully-featured API allows users to interact with Tribe programmatically if desired. Tribe implements the OAuth 2.0 standard as well as gene identifier mapping, which facilitates its integration into existing servers. Access to Tribe’s resources is facilitated by an easy-to-install Python application called tribe-client. We provide Tribe and tribe-client under a permissive open-source license to encourage others to download the source code and set up a local instance or to extend its capabilities.ConclusionsThe Tribe webserver addresses challenges that have made reproducible multi-webserver workflows difficult to implement until now. It is open source, has a user-friendly web interface, and provides a means for researchers to perform reproducible gene set based analyses seamlessly across webservers and command line tools.

One way to design a drug is to attempt to phenocopy a genetic variant that is known to have the desired effect. In general, drugs that are supported by genetic associations progress further in the development pipeline. However, the number of associations that are candidates for development into drugs is limited because many associations are in noncoding regions or difficult to target genes. Approaches that overlay information from pathway databases or biological networks can expand the potential target list. In cases where the initial variant is not targetable or there is no variant with the desired effect, this may reveal new means to target a disease. In this review we discuss recent examples in the domain of pathway and network-based drug repositioning from genetic associations. We highlight important caveats and challenges for the field, and we discuss opportunities for further development.

Central to the U.S. precision medicine initiative is the notion that Big Data, properly collected and analyzed, will reveal subtypes of diseases we currently view as single disorders. But what about small data sets? Is there translational value to meta-analyses of existing or newly generated data sets that aren’t part of large studies? Now, Planey and Gevaert bring together small data sets and show that the answer is yes. The investigators aim was to identify subtypes of disease, but many of the studies they used were too small to reveal robust subtypes on their own. So they developed a computational meta-analysis procedure called CoINcIDE—which used clustering to identify disease subtypes within each individual data set—and found cross–data set clusters of similar subtypes. Performing clustering within studies followed by cross-study comparison sidesteps obstacles such as platform or study effects that could hinder subtype discovery. Cross–data set analysis with CoINcIDE identified robust disease subtypes in both simulated and patient biopsy data from studies of breast and ovarian cancer. The authors performed the analysis of ovarian cancers both with and without the largest data set, TCGA. With 578 ovarian cancers, TCGA was more than twice as large as any other data set. Results were consistent whether or not TCGA was included, highlighting the power of smaller data sets paired with the CoINcIDE method for subtype discovery. The CoINcIDE analyses add to the literature on ovarian-cancer subtypes associated with survival; however, work to confirm the predictive nature of these subtypes or to identify effective subtype-specific treatments remains to be done. As a next step, the biological basis of the gene expression signals that give rise to the CoINcIDE subtypes needs to be identified, and the underlying mechanisms should be determined. Vincent van Gogh wrote, “Great things are not done by impulse, but by a series of small things brought together.” CoINcIDE might empower researchers to perform subtype-discovery for rare diseases, which otherwise would be impossible for a single lab or initiative because of the small numbers of samples. By developing a computational method that brings together small data sets, the authors transform the vast array of disease-relevant, publicly available data into opportunities for subtype discovery. C. R. Planey, O. Gevaert, CoINcIDE: A framework for discovery of patient subtypes across multiple data sets. Genome Med . 10.1186/s13073-016-0281-4 (2016). [[Full Text]][1] [1]: https://genomemedicine.biomedcentral.com/articles/10.1186/s13073-016-0281-4

The development of technology capable of inexpensively performing large-scale measurements of biological systems has generated a wealth of data. Integrative analysis of these data holds the promise of uncovering gene function, regulation, and, in the longer run, understanding complex disease. However, their analysis has proved very challenging, as it is difficult to quickly and effectively assess the relevance and accuracy of these data for individual biological questions. Here, we identify biases that present challenges for the assessment of functional genomics data and methods. We then discuss evaluation methods that, taken together, begin to address these issues. We also argue that the funding of systematic data-driven experiments and of high-quality curation efforts will further improve evaluation metrics so that they more-accurately assess functional genomics data and methods. Such metrics will allow researchers in the field of functional genomics to continue to answer important biological questions in a data-driven manner.

Deep learning describes a class of machine learning algorithms that are capable of combining raw inputs into layers of intermediate features. These algorithms have recently shown impressive results across a variety of domains. Biology and medicine are data-rich disciplines, but the data are complex and often ill-understood. Hence, deep learning techniques may be particularly well suited to solve problems of these fields. We examine applications of deep learning to a variety of biomedical problems—patient classification, fundamental biological processes and treatment of patients—and discuss whether deep learning will be able to transform these tasks or if the biomedical sphere poses unique challenges. Following from an extensive literature review, we find that deep learning has yet to revolutionize biomedicine or definitively resolve any of the most pressing challenges in the field, but promising advances have been made on the prior state of the art. Even though improvements over previous baselines have been modest in general, the recent progress indicates that deep learning methods will provide valuable means for speeding up or aiding human investigation. Though progress has been made linking a specific neural network's prediction to input features, understanding how users should interpret these models to make testable hypotheses about the system under study remains an open challenge. Furthermore, the limited amount of labelled data for training presents problems in some domains, as do legal and privacy constraints on work with sensitive health records. Nonetheless, we foresee deep learning enabling changes at both bench and bedside with the potential to transform several areas of biology and medicine.

The development of rapid data-collection technologies is changing the biomedical and biological sciences. In human genetics chip-based methods facilitate the measurement of thousands of DNA sequence variations from across the human genome. The collection of genetic data is no longer a major rate limiting step. Instead the new challenges are the analysis and interpretation of these high dimensional and frequently noisy datasets. The specific challenge we are interested in is the identification of combinations of interacting DNA sequence variations predictive of common human diseases. Specifically, we wish to detect epistasis or gene-gene interactions. Here we focus solely on the situation where there is an epistatic effect but no detectable main effect. The challenge for applying search algorithms to this problem is that the accuracy of a model is not indicative of the quality of the attributes within the model. Instead we use pre-processing of the dataset to provide building blocks which enable our evolutionary computation strategy to discover an optimal model.

Patient interactions with health care providers result in entries to electronic health records (EHRs). EHRs were built for clinical and billing purposes but contain many data points about an individual. Mining these records provides opportunities to extract electronic phenotypes, which can be paired with genetic data to identify genes underlying common human diseases. This task remains challenging: high quality phenotyping is costly and requires physician review; many fields in the records are sparsely filled; and our definitions of diseases are continuing to improve over time. Here we develop and evaluate a semi-supervised learning method for EHR phenotype extraction using denoising autoencoders for phenotype stratification. By combining denoising autoencoders with random forests we find classification improvements across multiple simulation models and improved survival prediction in ALS clinical trial data. This is particularly evident in cases where only a small number of patients have high quality phenotypes, a common scenario in EHR-based research. Denoising autoencoders perform dimensionality reduction enabling visualization and clustering for the discovery of new subtypes of disease. This method represents a promising approach to clarify disease subtypes and improve genotype-phenotype association studies that leverage EHRs.GRAPHICAL ABSTRACTHIGHLIGHTSDenoising autoencoders (DAs) can model electronic health records.Semi-supervised learning with DAs improves ALS patient survival predictions.DAs improve patient cluster visualization through dimensionality reduction.

BACKGROUND: A major bottleneck in our understanding of the molecular underpinnings of life is the assignment of function to proteins. While molecular experiments provide the most reliable annotation of proteins, their relatively low throughput and restricted purview have led to an increasing role for computational function prediction. However, assessing methods for protein function prediction and tracking progress in the field remain challenging. RESULTS: We conducted the second critical assessment of functional annotation (CAFA), a timed challenge to assess computational methods that automatically assign protein function. We evaluated 126 methods from 56 research groups for their ability to predict biological functions using Gene Ontology and gene-disease associations using Human Phenotype Ontology on a set of 3681 proteins from 18 species. CAFA2 featured expanded analysis compared with CAFA1, with regards to data set size, variety, and assessment metrics. To review progress in the field, the analysis compared the best methods from CAFA1 to those of CAFA2. CONCLUSIONS: The top-performing methods in CAFA2 outperformed those from CAFA1. This increased accuracy can be attributed to a combination of the growing number of experimental annotations and improved methods for function prediction. The assessment also revealed that the definition of top-performing algorithms is ontology specific, that different performance metrics can be used to probe the nature of accurate predictions, and the relative diversity of predictions in the biological process and human phenotype ontologies. While there was methodological improvement between CAFA1 and CAFA2, the interpretation of results and usefulness of individual methods remain context-dependent., We acknowledge the contributions of Maximilian Hecht, Alexander Grün, Julia Krumhoff, My Nguyen Ly, Jonathan Boidol, Rene Schoeffel, Yann Spöri, Jessika Binder, Christoph Hamm and Karolina Worf. This work was partially supported by the following grants: National Science Foundation grants DBI-1458477 (PR), DBI-1458443 (SDM), DBI-1458390 (CSG), DBI-1458359 (IF), IIS-1319551 (DK), DBI-1262189 (DK), and DBI-1149224 (JC); National Institutes of Health grants R01GM093123 (JC), R01GM097528 (DK), R01GM076990 (PP), R01GM071749 (SEB), R01LM009722 (SDM), and UL1TR000423 (SDM); the National Natural Science Foundation of China grants 3147124 (WT) and 91231116 (WT); the National Basic Research Program of China grant 2012CB316505 (WT); NSERC grant RGPIN 371348-11 (PP); FP7 infrastructure project TransPLANT Award 283496 (ADJvD); Microsoft Research/FAPESP grant 2009/53161-6 and FAPESP fellowship 2010/50491-1 (DCAeS); Biotechnology and Biological Sciences Research Council grants BB/L020505/1 (DTJ), BB/F020481/1 (MJES), BB/K004131/1 (AP), BB/F00964X/1 (AP), and BB/L018241/1 (CD); the Spanish Ministry of Economics and Competitiveness grant BIO2012-40205 (MT); KU Leuven CoE PFV/10/016 SymBioSys (YM); the Newton International Fellowship Scheme of the Royal Society grant NF080750 (TN). CSG was supported in part by the Gordon and Betty Moore Foundation’s Data-Driven Discovery Initiative grant GBMF4552. Computational resources were provided by CSC – IT Center for Science Ltd., Espoo, Finland (TS). This work was supported by the Academy of Finland (TS). RCL and ANM were supported by British Heart Foundation grant RG/13/5/30112. PD, RCL, and REF were supported by Parkinson’s UK grant G-1307, the Alexander von Humboldt Foundation through the German Federal Ministry for Education and Research, Ernst Ludwig Ehrlich Studienwerk, and the Ministry of Education, Science and Technological Development of the Republic of Serbia grant 173001. This work was a Technology Development effort for ENIGMA – Ecosystems and Networks Integrated with Genes and Molecular Assemblies (http://enigma.lbl.gov), a Scientific Focus Area Program at Lawrence Berkeley National Laboratory, which is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Biological & Environmental Research grant DE-AC02-05CH11231. ENIGMA only covers the application of this work to microbial proteins. NSF DBI-0965616 and Australian Research Council grant DP150101550 (KMV). NSF DBI-0965768 (ABH). NIH T15 LM00945102 (training grant for CSF). FP7 FET grant MAESTRA ICT-2013-612944 and FP7 REGPOT grant InnoMol (FS). NIH R01 GM60595 (PCB). University of Padova grants CPDA138081/13 (ST) and GRIC13AAI9 (EL). Swiss National Science Foundation grant 150654 and UK BBSRC grant BB/M015009/1 (COD). PRB2 IPT13/0001 - ISCIII-SGEFI / FEDER (JMF)., This is the final version of the article. It first appeared from BioMed Central at http://dx.doi.org/10.1186/s13059-016-1037-6.

Large, publicly available gene expression datasets are often analyzed with the aid of machine learning algorithms. Although RNA-seq is increasingly the technology of choice, a wealth of expression data already exist in the form of microarray data. If machine learning models built from legacy data can be applied to RNA-seq data, larger, more diverse training datasets can be created and validation can be performed on newly generated data. We developed Training Distribution Matching (TDM), which transforms RNA-seq data for use with models constructed from legacy platforms. We evaluated TDM, as well as quantile normalization, nonparanormal transformation, and a simplelog2transformation, on both simulated and biological datasets of gene expression. Our evaluation included both supervised and unsupervised machine learning approaches. We found that TDM exhibited consistently strong performance across settings and that quantile normalization also performed well in many circumstances. We also provide a TDM package for the R programming language.

Big data bring new opportunities for methods that efficiently summarize and automatically extract knowledge from such compendia. While both supervised learning algorithms and unsupervised clustering algorithms have been successfully applied to biological data, they are either dependent on known biology or limited to discerning the most significant signals in the data. Here we present denoising autoencoders (DAs), which employ a data-defined learning objective independent of known biology, as a method to identify and extract complex patterns from genomic data. We evaluate the performance of DAs by applying them to a large collection of breast cancer gene expression data. Results show that DAs successfully construct features that contain both clinical and molecular information. There are features that represent tumor or normal samples, estrogen receptor (ER) status, and molecular subtypes. Features constructed by the autoencoder generalize to an independent dataset collected using a distinct experimental platform. By integrating data from ENCODE for feature interpretation, we discover a feature representing ER status through association with key transcription factors in breast cancer. We also identify a feature highly predictive of patient survival and it is enriched by FOXM1 signaling pathway. The features constructed by DAs are often bimodally distributed with one peak near zero and another near one, which facilitates discretization. In summary, we demonstrate that DAs effectively extract key biological principles from gene expression data and summarize them into constructed features with convenient properties.

The genetic basis for primary open-angle glaucoma (POAG) is not yet understood but is likely the result of many interacting genetic variants that influence risk in the context of our local ecology. The complexity of the genotype to phenotype mapping relationship for common diseases like POAG necessitates analytical approaches that move beyond parametric statistical methods such as logistic regression that assume a particular mathematical model. This is particularly important in the era of big data where it is routine to collect and analyze data sets with hundreds of thousands of measured genetic variants in thousands of human subjects. We introduce here the Exploratory Modeling for Extracting Relationships using Genetic and Evolutionary Navigation Techniques (EMERGENT) algorithm as an artificial intelligence approach to the genetic analysis of common human diseases. EMERGENT builds models of genetic variation from lists of mathematical functions using a form of genetic programming called computational evolution. A key feature of the system is the ability to utilize pre-processed expert knowledge giving it the ability to explore model space much as a human would. We describe this system in detail and then apply it to the genetic analysis of POAG in the Glaucoma Gene Environment Initiative (GLAUGEN) study that included approximately 1,272 subjects with the disease and 1057 healthy controls. A total of 657,366 single-nucleotide polymorphisms (SNPs) from across the human genome were measured in these subjects and available for analysis. Analysis using the EMERGENT framework revealed a best model consisting of six SNPs that map to at least six different genes. Two of these genes have previously been associated with POAG in several studies. The others represent new hypotheses about the genetic basis of POAG. All of the SNPs are involved in non-additive gene-gene interactions. Further, the six genes are all directly or indirectly related through biological interactions to the vascular endothelial growth factor (VEGF) gene that is an actively investigated drug target for POAG. This study demonstrates the routine application of an artificial intelligence-based system for the genetic analysis of complex human diseases.

We present SEEK (search-based exploration of expression compendia; http://seek.princeton.edu/), a query-based search engine for very large transcriptomic data collections, including thousands of human data sets from many different microarray and high-throughput sequencing platforms. SEEK uses a query-level cross-validation-based algorithm to automatically prioritize data sets relevant to the query and a robust search approach to identify genes, pathways and processes co-regulated with the query. SEEK provides multigene query searching with iterative metadata-based search refinement and extensive visualization-based analysis options.

The biggest challenge for text and data mining is to truly impact the biomedical discovery process, enabling scientists to generate novel hypothesis to address the most crucial questions. Among a number of worthy submissions, we have selected six papers that exemplify advances in text and data mining methods that have a demonstrated impact on a wide range of applications. Work presented in this session includes data mining techniques applied to the discovery of 3-way genetic interactions and to the analysis of genetic data in the context of electronic medical records (EMRs), as well as an integrative approach that combines data from genetic (SNP) and transcriptomic (microarray) sources for clinical prediction. Text mining advances include a classification method to determine whether a published article contains pharmacological experiments relevant to drug-drug interactions, a fine-grained text mining approach for detecting the catalytic sites in proteins in the biomedical literature, and a method for automatically extending a taxonomy of health-related terms to integrate consumer-friendly synonyms for medical terminologies.