Combining machine learning and population genetics to elucidate evolutionary histories

Invited Speaker

Prof. Dan Schrider

Description

Genome sequence data can be used to answer a host of questions about evolutionary history. To what extent is genetic variation shaped by natural selection and recombination? Has a population experienced substantial changes in size? To what extent has there been gene flow between populations? In recent decades theoretical and methodological advances in population genetics have sought to address each of these questions. In this talk I will describe my efforts adapting supervised machine learning methods to answer evolutionary questions. I will demonstrate that these approaches outperform more traditional modes of population genetic inference, and yield valuable insights about recent evolutionary history including the process of recent and ongoing adaptation in human populations.