The AA and EA sample sizes provided ∼90% power to detect variants with a MAF ≥ 0.1% and nearly 100% power to detect common variants (MAF ≥ 5%) (tables S1 and S2 and fig. S10). With our large sample size, the proportion of singletons identified rapidly decreased to a nearly constant rate of new singleton discovery such that each additional exome contributed ∼200 novel SNVs (fig. S11). The number of SNVs per gene rapidly stabilized for common variants in small sample sizes (∼100 individuals), whereas the number of rare variants continued to increase linearly (fig. S12).
