Replication of associations previously reported in Caucasians in the CARe African-American meta-analyses.

To be included in this table, we require a two-tailed P≤0.05 after Bonferroni correction for the number of independent loci reported in the NHGRI database (Table S5). We report the association results for the published (index) SNP, unless it is not available. In that case, we report results for a proxy SNP (r2≥0.5 with original SNP in HapMap CEU; see Table S5 for additional details).
aProxy SNPs are marked with (p). SNPs that have a strong association signal but are not in LD with the published SNP are marked with (n) as potentially novel.
bPosition on NCBI build 36.1.
cEffect alleles are given on the forward strand. For proxy SNPs, we phased HapMap CEU genotypes for the index and proxy SNPs to determine haplotypes, to be able to assess consistency of the direction of effect.
dFor dichotomous phenotypes, we report odds ratio (OR) and 95% confidence interval (CI); for quantitative traits, we report effect size (beta, in standard deviation units) and standard error (SE).
eP-values are corrected using genomic control.