Speaker
Description
Suppose that mutant invasion starts with $k$ mutants distributed over a population structure. Which $k$ locations should we choose for the initial mutants so as to maximize their fixation probability?
In another setting, suppose that selection pressure is only present on a few locations in a population structure, therefore mutants have an advantage only when reproducing on these locations. Suppose, also, that we can control which locations feel the selection. Alas, we cannot choose all locations, but only $k$ of them. Which $k$ locations should we choose so as to maximize the fixation probability of a randomly arising mutant?
Optimization questions like the above are standard in popular stochastic diffusion models such as in SIR and IC/IM, but have not been studied in the context of EGT. In this talk, I will present some of our early steps in this direction, including exciting results we proved, and annoying questions that persisted our efforts.
