Fig. 1: Overview of the ancIBD algorithm.
From: Accurate detection of identity-by-descent segments in human ancient DNA
a, Sketch of the ancIBD HHM. The HMM has five states: one background state of no allele sharing and four states modelling the four possible IBD-sharing states between two phased diploid genomes. We model phase switch errors within a true IBD segment as a transition between the four IBD states. b, Visualization of the full pipeline to call IBD. First, aDNA data are imputed and phased using GLIMPSE and a panel of modern reference haplotypes. We note that users can customize these upstream steps; for example, use other tools to obtain genotype likelihoods or use different reference panels. Our core software (ancIBD) is then applied to the imputed data to screen for IBD. It produces two tables, one listing all inferred IBD segments and one listing IBD summary statistics for each pair of individuals.
