View Program - 2022 Population, Evolutionary, and Quantitative Genetics Conference

The genetic basis of adaptation and its predictability are poorly understood. Examples of adaptation where the genetic basis has been mapped implicate selective sweeps at a few large-effect loci, despite our expectation that polygenic adaptation, which occurs via allele frequency changes at many small-effect loci, should be ubiquitous. Theoretical modeling of adaptation has been limited to only a few extreme settings (e.g., highly polygenic or Mendelian), and we do not have the theoretical basis to understand how adaptation behaves under other conditions. Here, we address this by modeling adaptation after a sudden shift in fitness optimum of a quantitative trait under a variety of settings, where we vary 1) the shift size, 2) the trait’s genetic complexity – ranging from the Mendelian to the highly polygenic extreme, and 3) the distribution of fitness effect, e.g., the proportion of large-effect alleles. We then describe both the phenotypic response to selection (changes in trait mean and variance over time) and the allelic response (e.g., which alleles contribute to adaptation). Importantly, we find sweep-like behavior is common when polygenicity is low, large effect alleles are common, and shift sizes are large relative to the heritable variance; otherwise, polygenic adaptative responses are the norm. We derive quantitative conditions on model parameters that predict different kinds of phenotypic and genetic responses. Our results help to bridge the gap between theory and known examples of adaptation, and inform efforts to identify the signals of adaptation in humans and other species. Additionally, the conditions for selective sweeps overlap with the conditions for parallel evolution, so our results also bear on understanding the predictability of adaptation.

When should we expect adaptation via a highly polygenic response vs selective sweeps?