Towards Inverse Modeling of Intratumoral Heterogeneity (1404.2724v3)

Abstract: Development of resistance limits efficiency of present anticancer therapies and preventing it remains big challenge in cancer research. It is accepted, at intuitive level, that the resistance emerges as a consequence of cancer cells heterogeneity at molecular, genetic and cellular levels. Produced by many sources, tumor heterogeneity is extremely complex time dependent statistical characteristics which may be quantified by the measures defined in many different ways, most of them coming from statistical mechanics. In the paper we apply Markovian framework to relate population heterogeneity with the statistics of environment. As, from the evolutionary viewpoint, therapy corresponds to a purposeful modification of the cells fitness landscape, we assume that understanding general relation between spatiotemporal statistics of tumor microenvironment and intratumor heterogeneity enables to conceive the therapy as the inverse problem and solve it by optimization techniques. To account for the inherent stochasticity of biological processes at cellular scale, the generalized distance-based concept was applied to express distances between probabilistically described cell states and environmental conditions, respectively.