Encircling the regions of the pharmacogenomic landscape that determine drug response

• DOI: 10.1186/s13073-019-0626-x
• Publication Date: 2019-03-26T14:02:58Z
• Authors: Adrià Fernández-T...; Miquel Duran-Frigola; Patrick Aloy
• Keywords: Pharmacogenomics; Drug response; Human genetics

The integration of large-scale drug sensitivity screens and genome-wide experiments is changing the field pharmacogenomics, revealing molecular determinants response without need for previous knowledge about action. In particular, transcriptional signatures may guide repositioning, prioritize combinations, point to new therapeutic biomarkers. However, inherent complexity signatures, with thousands differentially expressed genes, makes them hard interpret, thus giving poor mechanistic insights hampering translation clinics. To simplify we have developed a network-based methodology identify functionally coherent gene modules. Our strategy starts calculation drug-gene correlations followed by pathway-oriented filtering network-diffusion analysis across interactome. We apply our approach 189 drugs tested in 671 cancer cell lines observe connection between expression levels modules mechanisms action drugs. Further, characterize multiple aspects modules, including their functional categories, tissue-specificity, prevalence Finally, prove predictive capability demonstrate how they can be used as sets conventional enrichment analyses. Network biology strategies like module detection are able digest outcome pharmacogenomic initiatives, thereby contributing interpretability improving characterization screened.