The pitfalls and virtues of population genetic summary statistics: Detecting selective sweeps in recent divergences

Abstract

During evolution, genomes are shaped by a plethora of forces that can leave characteristic signatures. A common goal when studying diverging populations is to detect the signatures of selective sweeps, which can be rather difficult in complex demographic scenarios, such as under secondary contact. Moreover, the detection of selective sweeps, especially in whole‐genome data, often relies heavily on a narrow set of summary statistics that are affected by a multitude of factors, frequently leading to false positives and false negatives. Simulating genomic regions makes it possible to control these demographic and population genetic factors. We used simulations of large genomic regions to determine how different secondary contact and sympatric speciation scenarios affect the footprint of hard and soft selective sweeps in the presence of varying degrees of gene flow and recombination. We explored the ability of an array of population genetic summary statistics to detect the footprints of these selective sweeps under specific demographies. We focussed on metrics that do not require phased data or ancestral sequences and therefore have wide applicability. We found that a newly developed beta diversity measure, Bfalse¯GD utperformed all other metrics in detecting selective sweeps and that FST also performed well. High accuracy was also found in Δπ and genotype distance‐derived metrics. The performance of most metrics strongly depended on factors such as the presence of an allopatric phase, migration rates, recombination, population growth, and whether the sweep was hard or soft. We provide suggestions for locating and analysing the response to selective sweeps in whole‐genome data.

Divergence of olfactory receptors associated with the evolution of assortative mating and reproductive isolation in mice

“…Given the highly polygenic nature of genetic architecture underlying the behavioural shift and the apparent strong linkage disequilibrium among genes involved in this phenotypic divergence, candidate loci are likely under polygenic selection, expected for such complex traits (Latta 1998;Sella and Barton 2019). The fact that we did not find signatures of selective sweeps using SFSbased methods in most candidate outlier regions may be another indication of polygenic adaptation, known to be more detectable with pairwise tests for allelic differentiation (Stölting et al 2015;Schneider et al 2021). This result may also suggest that those candidate regions are more likely subject to soft sweeps than hard sweeps (Wang et al unpublished data).…”

Capturing the rapidly evolving study of adaptation

“…They find that, among 16 summary statistics, a recently developed summary statistic and a more traditional F ST performed best. Importantly, by simulating several conditions, Schneider et al (2021) show that different summary statistics perform best in different conditions. Selveshwari et al (2021), in contrast, utilized a combination of experimental evolution and high-throughput sequencing approaches.…”

“…For instance, Schneider et al. (2021) conducted a simulation study to explore how well diverse statistics can detect selective sweeps in recently diverged populations. They find that, among 16 summary statistics, a recently developed summary statistic and a more traditional F ST performed best.…”

“…Importantly, by simulating several conditions, Schneider et al. (2021) show that different summary statistics perform best in different conditions. Selveshwari et al.…”

“…This special issue aims to crystallize, in journal article form, a diverse and exciting set of methodological novelties and empirical findings. For instance, Schneider et al (2021) conducted a simulation study to explore how well diverse statistics can detect selective sweeps in recently diverged populations. They find that, among 16 summary statistics, a recently developed summary statistic and a more traditional F ST performed best.…”