Statistical and mathematical challenges in coalescent modeling of biological data

Statistical inference in population genetics heavily relies on coalescent models and have been successfully applied for the last 20 years. However, these models are not readily portable to understanding cell evolution and cancer tumor evolution. In this talk, I will present an overview of mathematical models of cancer tumor evolution and the limitations of these models for statistical inference of meaningful parameters from available data. I will then describe how coalescent modeling can provide an efficient alternative for inference. I will present some challenges, pose some questions and propose a solution for hypotheses testing.