A complete list of the 127 SNPs used in this study, along with the results of the univariate RPM analyses on the hypothesis generating data (COGEND) can be found in Table S1 (supplementary information). The phenotype was case–control status for nicotine dependence. The table lists the nicotine receptor gene cluster associated with each SNP, the chromosome, the base-pair location, the proportion of variance attributable to the SNP, and an empirical p value for association. The permutation-based empirical p values resulted from a maximum of 100,000 permutations of the phenotype for each SNP. The predictors were the SNPs treated as 3-level categorical variables. The results are consistent with the primary analyses (univariate) from the first-stage analysis of the COGEND data using logistic regression methods (Bierut et al. 2007; Saccone et al. 2009a).
