The majority of genotyping was conducted in Dr. David Goldman’s laboratory at NIAAA using the Ilumina GoldenGate method. For details of study design, see Hodgkinson and colleagues (2008). In instances where additional SNPs had to be genotyped to complete tagging in our sample, we selected LD-tagging SNPs (tSNPs) with Tagger (de Bakker et al., 2005) as implemented in Haploview 3.2 (Barrett et al., 2005) using the default criteria of r2 ≥ 0.8 and minor allele frequency (MAF) ≥0.2. As common variation is generally considered to be ≥0.05, our tagging SNPs capture very common variation. Because SNPs genotyped using the Ilumina GoldenGate platform were chosen based on being African haplotype tagging, some SNPs have a MAF < 0.2 in our Caucasian sample. However, no SNP has an MAF < 0.01, which was our threshold for eliminating SNPs in the overall sample. For genes displaying several isoforms, the longest isoform was chosen for tag selection but to limit genotyping load and cost, 5′ and 3′ regions of the genes and ESTs were not directly tagged.
