Identify signatures of selection and introgression

Written by: Keaka Farleigh, Ph.D.
Date: September, 21st, 2026.
Date last modified: September, 21st, 2026

Purpose

To help you understand how to identify signatures of selection driving divergence and evidence of balancing selection and introgression using ancestral recombination graphs (ARG) in ARGHelpR.

Background

ARGHelpR provides a function to identify evidence of selection associated with divergence (identify_argcandidates_divergence) and a function to identify evidence of balancing selection and introgression (identify_argcandidates_shared). Note that you need to pair ARG statistics with other tests, because distinct processes can lead to similar signatures. This is why we provide two functions, the identify_argcandidates_divergence is meant to be used in a case where there is a differentiation peak or similar and the identify_argcandidates_shared is meant to be used when you know there is shared polymorphism.

We provide a visualization below to help you understand what tests ARGHelpR is performing when you use these functions (see figure below).

Figure 1. Cartoons showing the expected relationship between ancestral recombination graphs that are driven by one of a variety of processes (colored phylogeny) relative to the genomic background (gray phylogeny). The table below each comparison shows whether the time to the most recent common ancestory between populations (TMRCAB) and the time to the most recent common ancestor within populations (TMRCAW) is greater than (up arrow), equal to (equal sign), or less than (down arrow) the genomic background; a dash indicates a statistic isn’t considered. The tests available in identify_argcandidates_shared is shown on the top and the tests available in identify_argcandidates_divergence are shown on the bottom. Users can specify analysis = ‘all’ to test each scenario or the character string in parentheses for each test.