A key goal of evolutionary biology and human genetics is to understand the ways in which natural selection has shaped genetic and phenotypic variation within and among populations. The vast amount of population-genomic data, ancient DNA sequence data, and genotype-phenotype mapping data being generated bring an unprecedented power to address long-standing questions about the impact of adaptation in human evolution and the role of migration and selection in driving genome-wide evolutionary change. Our lab brings new tools from the intersection of population genetics and statistics to address these questions, capitalizing on a range of new approaches, from building genomic predictions of traits to the construction of genome-wide evolutionary genealogies. To maximize the potential of these new data and approaches, we propose to develop novel population-genomic models and statistical tools that address the roles of natural selection and population structure in shaping population-genomic variation. Specifically, the proposed work will: 1) Clarify how GWAS and population genetics can be leveraged to both robustly identify signals of polygenic adaptation and problems with confounding in GWAS; 2) Use genome-wide genealogies to estimate recent histories of dispersal and selection; and 3) Estimate the proportion of allele-frequency change driven by linked selection and the time scales over which selection acts, genome-wide. The results of these projects will address fundamental questions about the structure of human genomic variation.
Over the past two decades, vast amounts of genomic data have been generated in medical genomics to identify the genetic basis of human disease. This proposal describes a set of research projects that will help the field to understand the impact of environmental confounding in such studies and the evolutionary forces that shape human genomic and phenotypic variation.
