Fig. 2From: KIN: a method to infer relatedness from low-coverage ancient DNAComparison of pairwise difference data and inferred IBD fragments. The top panel shows the proportion of differences in each window along the genome for a pair of simulated siblings. Dashed lines represent \(p_0\), \(p_1\) and \(p_2\) estimates. The second panel shows the true IBD state for each window. The remaining panels show the IBD states predicted by particular relatedness models. The log-likelihood value for each model is shown on upper left corner of the panel. Light and shaded backgrounds represent distinct chromosomesBack to article page