There has been extensive research dedicated to identifying predictive biomarkers of tumor sensitivity to ligand-based therapies. Although some biomarkers have emerged, there is still no comprehensive understanding that unifies the observed tumor cell type-specific behaviors, in particular the differential responses to anti-cancer drugs.The collective realization is that no single protein measurement is predictive of therapeutic outcome. Large-scale studies focusing on multiple genetic biomarkers may be partially inadequate given the observed fractional killing of genetically-identical clonal cell populations.The objective of this proposal is to address this deficiency by coupling high-content analyses (dynamic live cell imaging, high-throughput assays) with theoretical and computational methods to provide a global understanding of the origins of tumor cell heterogeneity in response to anti-cancer drugs that target the apoptotic signaling pathway at the receptor level. Ultimately this approach will serve to identify, on a tumor cell type basis, the most predictive set of biomarkers and design optimal therapeutic combinations.