TABLE 2

Estimates for model log-likelihood, ancestry, and other parameters for the structure runs shown in Figure 4

Admixture modelLinkage model
H1H2Log P(X|K)αH1H2Log P(X|K)αr
A.
10.9990.999-43060.0260.9950.515-43040.02863.5
20.9950.512-54140.10420.9660.523-46940.18972.4
40.9910.533-57280.19440.9400.522-46760.37115.3
80.9900.400-59650.17850.9320.505-48720.43687.3
160.9900.419-59570.1910.9370.492-49610.435314.2
320.9930.221-61530.11440.9560.479-54280.402136.4
640.9900.169-59700.09850.9610.476-54330.388662.1
1280.9950.044-59350.0380.9620.450-57810.3977106.7
2560.9970.016-58360.03030.9950.026-58560.0345114.3
B.
11.0000.470-393620.02311.0000.471-392190.024739.3
20.9990.501-496020.09070.9980.501-423600.09182.5
40.9990.531-557700.17860.9960.472-438810.20323.9
80.9990.508-572840.160.9960.510-449380.21037.4
160.9990.490-572990.18950.9970.507-465420.204715.3
320.9990.468-569690.18930.9970.505-487550.196232.2
640.9990.452-565580.18790.9980.487-507870.185562.2
1280.9980.366-560270.19180.9980.443-528570.1805110.4
2560.9990.006-548320.03210.9980.327-543730.1784199.6
C.
10.9980.412-621750.030.9990.411-621410.0250.00
20.9890.493-640330.16690.9930.486-634910.11522.46
40.9870.641-661790.28270.9870.570-653630.26614.97
80.9900.683-661450.27690.9840.592-654640.30618.82
160.9910.667-662130.24590.9850.592-658780.301618.3
320.9890.598-656570.26410.9850.578-656890.309740.81
640.9820.453-654290.34250.9850.456-653430.304158.37
1280.9950.024-643700.0540.9970.014-643930.031497.46
D.
10.9930.526-585170.05790.5750.450-534286.25423
20.9760.488-609280.22560.5730.450-542678.65439
40.9790.639-619220.32780.5410.441-5457010.71373
80.9860.703-630550.30220.5560.455-5413010.79377
160.9870.634-622970.27080.6050.481-544199.62402
320.9700.566-622480.39240.6020.470-543248.63426
640.9800.551-613680.3480.6040.459-539019.1437
1280.9690.261-625900.32390.6210.440-530297.69465
  • The first column gives the number of generations after the admixture event, T2. H1 and H2 are the average estimated ancestry from ancestral population 1 for individuals in the two postadmixture populations (true values are 1.000 and 0.500). r is measured in generations. The highest log-likelihood in each row gives a heuristic guide to which model fits better.