Model-Based Inference In Speciation Genomics

We are beginning to understand genome-wide evolution during speciation because of our increased capacity to sample genome-wide polymorphisms across many individuals and populations. One of the possible effects of sampling the genome at thousands or millions of genetic variants is to draw attention to multi-locus models for variation across the genome, including hierarchical Bayesian models that explicitly incorporate the dependence and shared information among loci and individuals. Herein I describe multi-locus hierarchical Bayesian models that quantify genome-wide evolutionary change during speciation. Moreover, I discuss how to best allocate sequencing effort among individuals for optimal model-based inference. I then describe two empirical studies that use hierarchical Bayesian models to test hypotheses regarding the genetics and evolution of barriers to gene flow. Both studies indicate that barriers to gene flow are often polygenic and evolve as a consequence of divergent selection.