Fig 1.
The effect of increasing stringency with ancestral LD pruning.
From left to right, ancestry proportions are 0.1, 0.25, 0.5, 0.75 and 0.9. |r| cutoffs are: none (red), 1.0 (orange), 0.9 (yellow), 0.8 (green), 0.7 (dark blue), 0.6 (cyan), 0.5 (indigo), and 0.4 (violet). The solid line indicates the expectation for unbiased time estimation. All read data were simulated with ploidy = 1. True admixture time was drawn from a uniform (0, 2000) distribution.
Fig 2.
Time estimates and accuracy statistics for samples of varying ploidies.
From left to right, ancestry proportions are 0.1, 0.25, 0.5, 0.75 and 0.9. Each sample ploidy is represented by one point color with ploidy one (black), two (red), ten (blue) and twenty (green). From top to bottom, each row is the estimated time in generations, the proportion of sites where the true state is within the 95% credible interval, the width of the 95% credible interval, the mean posterior error, and the proportion of sites where the maximum likelihood estimate is equal to the true state.
Fig 3.
Accuracy of the HMM for samples of high ploidy.
The 95% credible interval (shaded blue region), and the posterior mean (red) contrasted with the true ancestry frequencies (black). Simulated data were generated with an admixture time of 1500 generations, an ancestry proportion of 0.2, a sample ploidy of 100, and a mean sequencing depth of 25.
Table 1.
Parameter estimation and LAI when admixture occurs at a constant rate, rather than in a single pulse.
Table 2.
Parameter estimation and LAI when a subset of loci experience natural selection in the admixed population.
Fig 4.
Admixture time estimates for simulated data consistent with variation present in modern European and African populations.
From left to right, m = 0.1, m = 0.25, m = 0.5, m = 0.75, m = 0.9. The top row is completely phased chromosomes and the bottom row is unphased diploid data.
Fig 5.
Local ancestry of inversion bearing chromosomes (red) compared with those of standard arrangement chromosomes (black) for the same chromosome arm.
Positions of inversion breakpoints, as reported in [63,67] are shown as vertical dashed lines.
Fig 6.
The relationship between the proportion of African ancestry proportion and local recombination rates in 100 ancestry informative SNP windows within the Raleigh, NC population (left).
The correlation between the proportion of African ancestry proportion and local recombination rates in 100 ancestry informative SNP windows in all populations assayed (right). Lines indicate the 95% confidence interval obtained via block bootstrap replicates (see Methods).
Fig 7.
The partial correlation between LA and latitude with correction for chromosome-wide ancestry proportions.
Sites for which the probability of the observed clinal relationship was less than 0.005 were retained as significant (red). Inversion breakpoints for inversions that are at polymorphic frequencies on this ancestry cline are shown as dotted blue lines.
Fig 8.
The mean African ancestry proportion across all populations on the ancestry cline for chromosome arms 2L, 2R, 3L, 3R, and X (top to bottom).
Local minima outlier loci are shown in red (see Methods).