Supplementary Table 3a for simulations with continental-scale stratification and Supplementary Table 4a for simulations with national-scale stratification), and near 1 in simulations with polygenicity (Supplementary Figures 1–5). Furthermore the qualitative appearance of the pattern of inflation as a function of LD Score was completely different in each set of simulations (Fig 1c,d). The observed correlations between FST and LD Score in all simulations were negligible (generally 10−5 to 10−4, see Supplementary Tables 3b and 4b). We note that in simulations with population stratification, the LD Score regression slope was slightly greater than zero on average (Supplementary Tables 3c, 4c), likely a result of linked selection. Nevertheless, the performance of the LD Score regression intercept was comparable to λGC, and so would be suitably conservative if used as a correction factor, despite the small bias in the slope.
