Abstract:
The central goal of population genetics is to infer the evolutionary history of a population from observed genetic variation. However, the myriad of evolutionary processes often leave ambiguous signatures, so it can be difficult to reconstruct the evolutionary past. Classical methods typically assume that sequence variation is shaped by neutral processes such as generation to generation sampling variance, i.e., genetic drift.
During the course of my doctoral work, I have undertaken several projects aimed at analyzing populations that are dominated not by neutral processes but by adaptive ones. First, the HIV population within an infected patient experiences strong selection due to
immune pressure, and a low recombination rate causes beneficial mutations to sweep concurrently and interfere. To this end, I constructed a realistic model for the evolution of HIV and a method for inferring the selection coefficients of beneficial mutations thereby. Second, many tests of natural selection fail to distinguish between demographic expansion and rapid adaptation. Therefore, I developed a novel method that quantifies the collective effect of many mutations in the genome. The method does not depend on assumptions about demography and can indicate whether genetic draft (i.e., widespread hitchhiking) or genetic drift is the major factor shaping neutral variation. Finally, qualitative differences between rapidly adapting and neutrally evolving asexual populations, such as the statistics of their genealogies, are increasingly well understood, so I contributed to a project that extends coalescence in asexual populations to sexual populations. Properties of sexual populations can therefore be reduced to those of asexual populations, with suitably rescaled parameters.
With this work, I have helped to further a broad and current research program that recognizes the critical role of linked selection, interference, and genetic draft in interpreting patterns of genetic diversity.