Predicting synergistic target and drug target combinations in a large-scale data-driven simulation of cancer cells.

2084 A plethora of targeted therapies are emerging in clinical and pre-clinical trials with varying degrees of efficacy and toxicities, often depending on the cell type or patient genetic background. One approach for circumventing lack of efficacy is to use therapies in combination. The challenge is to come up with combinations that are synergistic to cancer cells, but leave normal cells unharmed. We present a large-scale data-driven simulation of the genes and proteins controlling the signal transduction pathways that drive the cell cycle, the cell cycle machinery, and their connections to apoptosis. The simulation is data-driven meaning that it can be trained to reflect a variety of cell types and genetic backgrounds. This is crucial to being able to make predictions that are specific to certain cancer types and for determining universal properties that span cell types and various cancer genetic backgrounds. The simulation is used to both predict which combination of targets and drug targets that can synergize to lead to maximal growth inhibition and the biological mechanism that led to that synergy. A large-scale simulation confers the advantage of being able to quickly simulate a variety of possibilities both in terms of target combinations and various cell types and genetic backgrounds with minimal cost. Thus making it a very powerful tool to aid testing a large combinatorial number of possibilities when one considers target and drug target choices and various genetic backgrounds.