Genotyping, imputation and quality control have been described previously [36]. Briefly, genetic data were used to assign ancestry and families were classified as primarily European (EUR) or Admixed African American (AFR) ancestries according to the ancestry of the greatest proportion of family members [36]. Genotyping of 798 AFR individuals and 3270 EUR individuals included in the analytic sample was performed using the Illumina 2.5M array (Illumina, San Diego, CA, USA), the Illumina OmniExpress [37], the Illumina 1M array, or the Affymetrix Smokescreen array [38]. SNPs with a genotyping rate <98%, Hardy-Weinberg equilibrium violations (p<10−6), or with minor allele frequency (MAF) less than 3% were excluded from analyses. Mendelian inconsistencies were removed, after which data were imputed to 1000 genomes (Phase 3) using SHAPEIT [39] and IMPUTE2 [40]. Following imputation, dosage probabilities ≥ 0.90 were converted to hard calls. Mendelian errors in the imputed SNPs were reviewed and resolved as described previously [41, 42]. SNPs with an imputation information score < 0.30 or MAF < 0.03 were excluded from subsequent analysis.
