Abstract
Combination therapies have become a standard treatment of several complex diseases. High-throughput screening makes it possible to profile phenotypic effects of thousands of drug combinations in patient-derived cells and other pre-clinical model systems. However, due to the massive number of potential drug and dose combinations, large-scale multi-dose combinatorial screening requires extensive resources and instrumentation. This combinatorial explosion is especially daunting in anticancer drug treatment, where combinatorial response profiles are often highly variable even between individual patients with same cancer type. Therefore, personalized combination treatments targeting multiple cancer growth and survival pathways are required. This talk presents computational and experimental approaches to cost-effective drug combination prediction and testing that enable identification of customized synergistic combinations for individual cancer patients. Several case studies demonstrate how machine learning pinpoints cancer-selective synergies, hence reducing the likelihood of toxic effects, toward successful clinical applications.