PharmOmics: A Species- and Tissue-specific Drug Signature Database and Online Tool for Toxicity Prediction and Drug Repurposing
Yen-Wei Chen, Douglas Arneson, Graciel Diamente, Jennifer Garcia, Nima Zaghari, Paul Patel, Patrick Allard, and Xia Yang
Received Date: 28th October 19
Drug development has been hampered by a high failure rate in clinical trials due to suboptimal efficacy or safety issues that are not predicted by preclinical studies in model systems. A key contributor to this hurdle is our incomplete understanding of how drug molecules function across organ systems and species. Therefore, elucidating species- and tissue-specific molecular actions of drugs can provide systems level insights into therapeutic efficacy, toxicity, adverse drug reactions (ADRs), and between-species differences that are necessary for more effective translational medicine. Here, we present a comprehensive drug knowledgebase and analytical tool, PharmOmics, comprised of genomic footprints of drugs in individual tissues from multiple species (human, mouse, and rat) based on meta-analysis of global transcriptome data from public repositories (GEO, ArrayExpress, TG-GATEs, drugMatrix). Using multi-species multi-tissue gene expression signatures as molecular indicators of drug functions, we demonstrate the potential of PharmOmics to predict drugs that can be used for new disease indications or that may induce ADRs. PharmOmics was able to accurately predict known drugs for multiple diseases and identify potential drugs targeting non-alcoholic fatty liver disease for which no approved drugs are currently available. In addition, we found that drugs causing similar ADRs tend to have gene signatures with close proximity in gene regulatory networks. Comparison of our PharmOmics platform with other commonly used drug signature databases, most of which are based on data derived from in vitro cell lines, demonstrates that our tissue- and species-specific drug signatures serve as a complementary resource to support drug characterization.
Read in full at bioRxiv.
This is an abstract of a preprint hosted on an independent third party site. It has not been peer reviewed but is currently under consideration at Nature Communications.