DNA specimens in the Yale-Penn sample were genotyped on three microarrays: Yale-Penn 1 on the Illumina HumanOmni1-Quad v1.0 microarray (OMNI), Yale-Penn 2 on the Illumina Infinium Human Core Exome microarray (HCE), and Yale-Penn 3 on the Illumina Multi-ethnic Global Array (MEGA). QC of genotype data was performed as previously described [13]. Briefly, individuals and single nucleotide polymorphisms (SNPs) with a call rate < 98% and variants with minor allele frequency (MAF) < 1% were excluded prior to imputation. Pairwise identity-by-decent (IBD) was calculated with PLINK to determine genetic relatedness among individuals in the sample and individuals with a pairwise IBD estimate > 25% were assigned to the same family. Self-reported males with X chromosome heterozygosity > 20% and self-reported females with X chromosome heterozygosity < 20% were excluded. SNP genotype imputation was performed using the March 2012 1,000 Genomes reference panel (1,000 Genomes Project, 2012; http://www.1000genomes.org/) and IMPUTE2 [36] separately in AAs and EAs implemented on the Michigan imputation server (https://imputationserver.sph.umich.edu).
