Fig. 4.

(a,b) Two introgression (MSci) models and (c,d) two migration (IM) models used in simulation. The thin branches have the population size ${\theta }_0=0.002$ and the thick branches have ${\theta }_1=0.01$. In MSci model A, the species divergence/introgression times are ${\tau }_R=4{\theta }_0$, ${\tau }_S=3{\theta }_0$, ${\tau }_T=2{\theta }_0$, and ${\tau }_X={\tau }_Y=1.5{\theta }_0$. In MSci model B, ${\tau }_R=4{\theta }_0$, ${\tau }_S=3{\theta }_0$, ${\tau }_T={\theta }_0$, and ${\tau }_X={\tau }_Y=1.5{\theta }_0$. Introgression probability is $\phi =0.2$. In the IM model C, ${\tau }_R=4{\theta }_0$, ${\tau }_S=3{\theta }_0$, and ${\tau }_T=2{\theta }_0$, with migration occurring from species A to B over time period ($0,{\tau }_T$) at the rate $M=0.1$ migrants per generation. In the IM model D, ${\tau }_R=4{\theta }_0$, ${\tau }_S=3{\theta }_0$, and ${\tau }_T={\theta }_0$, with migration from species A to ST over time period (${\tau }_T,{\tau }_S$) at the rate $M=0.1$. (e) The 95% HPD CIs for parameters in 100 replicate datasets of $L=250$, 1,000, and 4,000 loci. Column labels refer to the simulation model followed by the analysis model; for example, “B-A” means that data were simulated under model B and analyzed under model A. The values of $\theta$ and $\tau$ parameters are multiplied by ${10}^3$. Black solid line indicates the true value.