For NIDA individuals allele frequencies for each SNP in each DNA pool were assessed based on hybridization to the 3–4 “perfect match” cells on each of three arrays, as described [14], [15]. We validated this approach (Fig. 1) [14], [15]. The intensities of the highest and lowest 5% of features on each array were monitored and the variances in signal between replicate hybridizations of DNA from each pool and between hybridization signals from pools of the same phenotype and ethnicity were assessed. For the detection of nominally positive SNPs we averaged the “perfect match” data for each SNP on each array, derived the arctangent of the ratio between hybridization intensities for A and B alleles, averaged the arctan A/B values for the three replicate arrays, divided the mean arctan A/B ratios for abusers by the mean arctan A/B ratios for controls to form an abuser/control ratio for each SNP, and generated a “t” statistic for the differences between arctan A/B in abusers and controls with corresponding p values (see Fig. S1 for this portion of the “cluster then converge” analysis and Fig S2 for initial quality control).
