Opportunities and obstacles for deep learning in biology and medicine.

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.

Causal modeling using network ensemble simulations of genetic and gene expression data predicts genes involved in rheumatoid arthritis.

Tumor necrosis factor α (TNF-α) is a key regulator of inflammation and rheumatoid arthritis (RA). TNF-α blocker therapies can be very effective for a substantial number of patients, but fail to work in one third of patients who show no or minimal response. It is therefore necessary to discover new molecular intervention points involved in TNF-α blocker treatment of rheumatoid arthritis patients. We describe a data analysis strategy for predicting gene expression measures that are critical for rheumatoid arthritis using a combination of comprehensive genotyping, whole blood gene expression profiles and the component clinical measures of the arthritis Disease Activity Score 28 (DAS28) score. Two separate network ensembles, each comprised of 1024 networks, were built from molecular measures from subjects before and 14 weeks after treatment with TNF-α blocker. The network ensemble built from pre-treated data captures TNF-α dependent mechanistic information, while the ensemble built from data collected under TNF-α blocker treatment captures TNF-α independent mechanisms. In silico simulations of targeted, personalized perturbations of gene expression measures from both network ensembles identify transcripts in three broad categories. Firstly, 22 transcripts are identified to have new roles in modulating the DAS28 score; secondly, there are 6 transcripts that could be alternative targets to TNF-α blocker therapies, including CD86--a component of the signaling axis targeted by Abatacept (CTLA4-Ig), and finally, 59 transcripts that are predicted to modulate the count of tender or swollen joints but not sufficiently enough to have a significant impact on DAS28.

TB database: an integrated platform for tuberculosis research.

The effective control of tuberculosis (TB) has been thwarted by the need for prolonged, complex and potentially toxic drug regimens, by reliance on an inefficient vaccine and by the absence of biomarkers of clinical status. The promise of the genomics era for TB control is substantial, but has been hindered by the lack of a central repository that collects and integrates genomic and experimental data about this organism in a way that can be readily accessed and analyzed. The Tuberculosis Database (TBDB) is an integrated database providing access to TB genomic data and resources, relevant to the discovery and development of TB drugs, vaccines and biomarkers. The current release of TBDB houses genome sequence data and annotations for 28 different Mycobacterium tuberculosis strains and related bacteria. TBDB stores pre- and post-publication gene-expression data from M. tuberculosis and its close relatives. TBDB currently hosts data for nearly 1500 public tuberculosis microarrays and 260 arrays for Streptomyces. In addition, TBDB provides access to a suite of comparative genomics and microarray analysis software. By bringing together M. tuberculosis genome annotation and gene-expression data with a suite of analysis tools, TBDB (http://www.tbdb.org/) provides a unique discovery platform for TB research.

Using ChemBank to probe chemical biology.

ChemBank (http://chembank.broad.harvard.edu/) is a public, Web-based informatics environment. ChemBank stores and makes freely available data derived from small molecules and small-molecule screens and has resources for relating and studying these data. Currently, ChemBank stores information on hundreds of thousands of small molecules and hundreds of biomedically relevant assays performed at the Broad Institute screening center. Web-based analysis tools are available within ChemBank to study the relationships between small molecules, cell measurements, and cell states. This unit demonstrates the use of ChemBank data to ask and answer questions relating to chemical biology and screening experiments contained within ChemBank.

Enabling a community to dissect an organism: overview of the Neurospora functional genomics project.

A consortium of investigators is engaged in a functional genomics project centered on the filamentous fungus Neurospora, with an eye to opening up the functional genomic analysis of all the filamentous fungi. The overall goal of the four interdependent projects in this effort is to accomplish functional genomics, annotation, and expression analyses of Neurospora crassa, a filamentous fungus that is an established model for the assemblage of over 250,000 species of non yeast fungi. Building from the completely sequenced 43-Mb Neurospora genome, Project 1 is pursuing the systematic disruption of genes through targeted gene replacements, phenotypic analysis of mutant strains, and their distribution to the scientific community at large. Project 2, through a primary focus in Annotation and Bioinformatics, has developed a platform for electronically capturing community feedback and data about the existing annotation, while building and maintaining a database to capture and display information about phenotypes. Oligonucleotide-based microarrays created in Project 3 are being used to collect baseline expression data for the nearly 11,000 distinguishable transcripts in Neurospora under various conditions of growth and development, and eventually to begin to analyze the global effects of loss of novel genes in strains created by Project 1. cDNA libraries generated in Project 4 document the overall complexity of expressed sequences in Neurospora, including alternative splicing alternative promoters and antisense transcripts. In addition, these studies have driven the assembly of an SNP map presently populated by nearly 300 markers that will greatly accelerate the positional cloning of genes.