Project 2 - Abstract. We view cancer as an open complex adaptive system that can be characterized using necessary and sufficient data, eco-evolutionary first principles and sophisticated computational methods. Here we examine perturbations of that system through therapy. We propose that optimization of treatment requires solid understanding of the underlying dynamics which, in turn, permits accurate predictions of its response to any therapeutic perturbation. . We will focus our efforts on three of these teams each of which will concentrate on active clinical trials that investigate different tumors (multiple myeloma, prostate cancer and glioblastoma) treated with different therapeutic strategies (multidrug chemotherapy, hormonal treatment, and immunotherapy). In each trial, the multidisciplinary team will apply an existing computational model that captures the intratumoral evolutionary and ecological dynamics using available clinical data. The general goal is to predict response and resistance to therapy in each patient. Using an iterative approach, each team will work with the B2B core and Project 1 to optimize the models' predictive power by exploring alternative mathematical methods and inclusion of new data elements from each trial (e.g. multi-parametric imaging, circulating tumor cells, etc.). In follow-on trials using the same tumors and treatment strategies, we will validate model predictions, apply variations in the treatment approaches (or combinations of approaches) suggested by parallel pre-clinical experiments in Project 1 and model simulations in the B2B core.