and adjusted for treatment. When the interaction was not significant, the data were analyzed as a single pooled sample, adjusted for treatment. To correct for multiple testing in which many markers were in strong linkage disequilibrium (LD), we selected individual SNPs representing bins of highly linked SNPs (r2>0.8), with a single SNP for each bin included in a false discovery rate, multiple test correction approach (bins were derived from the Caucasian population described here). This reduced the total estimated number of independent tests from 785 to 254. All of the p-values reported here fell below a q-value of 0.2, using this approach.
