Selection scans and downstream analysis with selscan

Abstract

Summary statistics based on Extended Haplotype Homozygosity (EHH) are widely used for inferring positive selection in genomes as a result of their ease of use, computational efficiency, and interpretability. These various summary statistics can be applied to single populations or to pairs of populations, can be used with a genetic recombination map or without, and can be applied to phased or unphased data. Although these statistics are straightforward to compute, there lacks clear descriptions on how they relate to one another, how they should be used, and how their resulting outputs should be interpreted. Here, we provide a comprehensive introduction to selection statistics as they are implemented in the widely used software, selscan. In addition to this detailed guide, we implement enhanced normalization procedures and support for gene-based analyses, enabling users to translate selection signals captured by these statistics into gene-level interpretations using BED annotation files, facilitating biologically meaningful insights. We demonstrate the behavior of such statistics on simulated data and highlight best practices by performing an example downstream analysis on data from the 1000 Genomes Project using new features in selscan v3.0. We hope these guidelines will foster reproducibility in the evolutionary genomics community. Precompiled executables and source code for selscan v3.0 can be found at https://github.com/szpiech/selscan.