The landscape of pharmacogenomic interactions in human cancer

Systematic studies of the cancer genome are providing unprecedented insights into the molecular nature of human cancer. Using this information to guide the development and application of therapies in the clinic remains challenging. Here we report how cancer driving alterations detected in 11,215 tumors and 29 different tissues (integrating multiple omics) correlate with response to 265 compounds profiled in 1,001 cancer cell lines. We find that cell lines faithfully resemble tumours on the domain of these alterations, and numerous examples of altered genes and pathways conferring drug sensitivity and resistance. Significantly, logic-based modeling improves our ability to identify drug sensitive sub-types and machine-learning methods enable us to investigate the predictive ability of the different data-omics. We provide a comprehensive analysis of how somatic alterations identified from the large-scale analysis of primary tumors impact on drug response. This represents a rich resource to help identify therapeutic options for selected cancer populations.