Quality control (QC) of the genotype data was performed before imputation. Subjects were excluded from analysis if they had <90% genotyping success or were population stratification outliers. Samples were evaluated for relatedness using identity by descent (IBD), and duplicate or related samples (IBD >0.25) were removed. Using AIMs and “null” SNPs, we performed principal component analysis (PCA, EIGENSTRAT [38], [39], and STRUCTURE [40] to identify outliers, hidden population structure, and estimate individual ancestry proportions as previously described [41]. SNPs were removed if they had (a) >10% missing genotypes, (b) showed evidence of deviation from Hardy-Weinberg equilibrium (PHWE<0.001 in controls or <0.0001 in cases), (c) for poor clustering, or had a minor allele frequency (MAF) <1%.
