Moffitt PSOC Here we focus on two deeply interconnected physical science questions: "How do we study, quantify, integrate, and model the complexity of cancer biology and treatment across multiple length and time scales that form the tumor ecology?" and "Can the evolutionary dynamics of therapeutic resistance be exploited through dynamic spatio-temporal models to optimize treatments and improve the lives of patients with cancer?" We propose that cancer must be investigated and treated as a complex adaptive systems in which the underlying first principles are Darwinian. We view intratumoral evolution as a dynamical interaction between environmental selection forces and tumor adaptive strategies to maximize fitness. A critical property of cancer complex system is that it is open andthus can be perturbed by host response and iatrogenic interventions. Thus, the multi-scale (e.g. molecular, cellular and tissue scales) spatio-temporal variations within and between cancers (i.e. the "ecology" of cancer) is dependent in large part on the open components of the system such as alterations in blood flow that affect local environmental conditions and subsequent cellular adaptive strategies. Similarly, the Darwinian response to therapy will vary within each habitat within the tumor ecology and must be understood to design consistently effective therapies. We approach these questions in two different ways: In project 1 we focus on fundamental principles - the cancer cell evolutionary dynamics and molecular mechanisms that permit adaptation to host-generated perturbations including blood flow and treatment strategies. Here the focus will be on sophisticated in-vitro and in-vivo experimental methods integrated with Darwinian-based mathematical models. A key deliverable from Project 1 is identification of novel therapeutic strategies that can exploit these evolutionary dynamics and molecular mechanism to improve clinical therapy. In Project 2 we will focus on developing computational models that use first principles and available clinical data to: 1. understand the patient-specific dynamics that govern response and resistance and 2. develop computational models that predict the outcomes of different therapies (e.g. multidrug chemotherapy, immunotherapy, and hormone therapy ) in individual patients. In the longer term our goal is to increase the scope of these models to permit design of patient-specific therapy to optimize overall survival. The deliverable from Project 2, therefore, include development of methods to extract maximum amounts of information from clinically available data and development of computational models to optimize clinical therapy using often sparse dynamic data. Both Projects will interact extensively with a core focused on developing computational models and applying sophisticated analytic methods to extract maximum knowledge from available molecular, pathological, and radiological clinical data.