Selection
Selection in the Kalash populations was estimated as described by Yi et al.30 In brief, we obtained estimates of the population time of divergence (T) from FST values and corrected for the effective population size of the population considered (Nep), as shown inT=log(1−FST)log(1−12×Nep).
For each marker, we calculated T between the Kalash and CHB (TKC), between the Kalash and Balochi (TKB)—who were considered representative of the other Pakistani populations—and between the CHB and Balochi (TCB). The population branch statistic (PBS), which defines the length of the branch leading to the Kalash since the split from East Asians, is equal toPBS=TKC+TKB−TCB2.
Nep estimates were obtained from the linkage data (2,471 [95% confidence interval (CI) = 2,319–2,603]) with the NeON R package. The variants within the 99th percentile of the distribution of our PBS values were annotated, and a list of genes associated with these variants was used for Ingenuity Pathway Analysis (IPA).