Fig. 2
From: A method for positive forensic identification of samples from extremely low-coverage sequence data
Summary of results from coalescent simulations. a Results from 100 replicates of comparisons between three low-coverage subsamples from two simulated diploid individuals with a reference panel of haplotypes drawn from the same population. The first three panels report the results of comparisons made with alleles observed within the same sample. The aggregated log-likelihood ratio values are largely positive, indicating that the observations are consistent with originating from a single diploid individual. Similarly, the fourth panel shows positive values for the comparison made between two independent subsamplings of the same diploid individual. In the last panel, comparisons of allele observations made from two independent individuals produce negative aggregated log-likelihood values, indicating the two samples are from different individuals. b A variation of the comparisons made in Panel a where samples from diploid individuals are compared using a panel of reference haplotypes from a related, but diverged population (see diagram). Each panel shows the results of five comparisons; four made between samples from the same individual (blue) and one made between different individuals (red). From these results, a closely related population retains power to differentiate between one or two individuals over many generations. As the number of generations since the population split increases, the power of the related population to model of the population from which the samples were drawn diminishes. c Results from a simulated model of recent human population history (see diagram and Additional file 1). In each subpopulation, comparisons between samples from diploid individuals using a panel of haplotypes and differentiate between the four comparisons made against the same individual and the one comparison made between different individuals. Increased linkage disequilibrium in subpopulations that have undergone recent population bottlenecks (Europe and Asia) increases the power of the method to differentiate between one and two individuals