It's time to stop sweeping recombination rate under the genome scan rug

Abstract

Different parts of the genome can vary widely in their evolutionary histories and sequence divergence from other species. Indeed, some of the most interesting biology (e.g., hybridization, horizontal gene transfer, variable mutation rates across the genome) is revealed by the discordant relationships between taxa across the genome. The goal for much of evolutionary genetics is centred on understanding the evolutionary processes by which such varied signatures arise and are maintained. Many evolutionary genetics studies seek to identify signatures of positive selection between two closely related ecotypes or taxa by delineating regions with particularly high divergence relative to a genome-wide average, often termed "divergence outliers." In a From the Cover article in this issue of Molecular Ecology, Booker et al. take a major step forward in showing that recombination rate differences are sufficient to create false positive divergence outliers, even under neutrality. They demonstrate that the variance of genome scan metrics is especially high in regions with low recombination rates, consistent with previous work. Furthermore, they show that both relative and absolute measures of divergence (F_ST and D_XY , respectively) as well as other commonly used statistics in genome scans (e.g., π_W , Tajima's D and H12) all have similar covariance between variance and local recombination rate. Finally, Booker et al. show that low recombination regions will tend to produce more outliers if genome-wide averages are used as cut-offs to define genomic outliers. Booker et al.'s results suggest that recombination rate variation, even under neutral conditions, can shape genome scans for selection, and this important variable can no longer be ignored.

Keywords: divergence outliers; islands of divergence; selection scans.