Skip to main content
. 2022 Feb 16;71(5):1159–1177. doi: 10.1093/sysbio/syac009

Figure 4.

Figure 4

a) The introgression (MSci) model, proposed based on the bpp species tree estimation under MSC and 3s analysis under the IM model, involving two unidirectional introgression events (s Inline graphic d with introgression probability Inline graphic, and tc1 Inline graphic c with probability Inline graphic and two BDI events (e Inline graphic h with probabilities Inline graphic and Inline graphic, and ds2 Inline graphic tc2 with probabilities Inline graphic and Inline graphic. The model involves 34 parameters: 5 species divergence times and 4 introgression times (Inline graphic), 19 population sizes (Inline graphic), and 6 introgression probabilities. To identify parameters (see Fig. S4 of the Supplementary material available on Dryad), internal nodes are labeled with lowercase letters, for example, e is the parent node of Era, eh is the parent node of e and h, etc. Each branch corresponds to a population and is referred to using the label for its daughter node, for example, branch e-Era is labeled branch Era and has population size Inline graphic, and branch eh-e is labeled branch e and has Inline graphic. Branch lengths in the tree represent posterior means of divergence/introgression times in the bpp analysis of the 6,030 noncoding loci in chromosome 1. Estimates for other chromosomes or for coding loci are presented in Figure 5. b) Posterior means (dots) and 95Inline graphic HPD intervals (bars) of the six introgression probabilities (Inline graphic) under the MSci model of (a) obtained from bpp analysis of the 25 chromosomal regions (see Figure 5 for the number of loci). Estimates for other parameters are in Figure S4 of the Supplementary material available on Dryad. Inversion regions 2b and 15b on chromosomes 2 and 15 were analyzed separately, and there was an alternative set of posterior estimates resulting from within-model unidentifiability; see Figure 6.