At each SNP allelic association was assessed between cases and control using the χ2 test with 1 d.f. Allelic odds ratio (OR) and 95% confidence intervals (95% CIs) were calculated using PLINK [43]. Haploview [44] was used to estimate the linkage disequilibrium (LD) between markers in each ethnic population. We performed pair-wise conditional haplotype analysis using WHAP [45] to identify which associations were independent, and not due to LD with primarily associated SNPs. For each SNP we assessed whether it was still associated with SLE after conditioning on the most significant SNP. To correct for multiple testing, false discovery rate (FDR) methods were used and q values were calculated using PLINK (43). Q values correspond to the proportion of false positives among the results. Thus, q values <0.05 signify less than a 5% false positive rate and are taken as a measure of significance.
