Genomic DNA was extracted from saliva [63], whole blood or buffy coat, quantified and normalized to 50ng/ul, and genotyped at the University of Southern California Epigenome Center, and at the University of California San Francisco Institute for Human Genetics Genomics Core Facility. We extracted SNP genotype data from custom 1536 SNP Illumina GoldenGate panels interrogating candidate genes of interest to PNAT [46,64] and imputed genotype data where necessary. All genotyping included HapMap and replicate DNA samples. We reviewed and filtered GoldenGate genotyping data as described [46] for RCTs 3A and 3B and in a similar fashion for the remaining RCTs by manual review of genotype cluster metrics, review of HapMap sample concordance, by successively filtering samples and SNPs with call rates below a defined threshold, and comparison of X chromosome heterozygosity and clinical gender. We estimated principal components of population genetic variation [65] among self-identified White participants using 45 ancestry informative markers genotyped across all individuals. Genotypes were imputed with IMPUTE v2.1.2 [66] using 1000 Genomes CEU (Utah residents with ancestry from northern and western Europe) August 2010 haplotype data
