We found GO terms for the cholinergic receptors, that included genes tagged by SNPs previously reported to achieve genome-wide significant levels (i.e. 5.0E−08), although these SNPs did not necessarily achieve this level in the datasets we evaluated. In addition, other cholinergic receptor genes with more modest p-values were also enriched in these GO terms. Each study identified CHRNA7 but the significant SNPs were different and in low r2 (<0.20) but high D’ (>0.80) values, suggesting the existence of a shared risk allele. However, we could not identify a same SNP that was significant for the three studies in the region (p-value<0.05). This might indicate the presence of an untyped SNP, possibly with a minor allele frequency too low to be accurately imputed, or might be a synthetic association representing the effects of multiple rare variants. In contrast, for the CHRNA9 gene, we could neither identify a common significant SNP nor a common allele tagged by SNPs in linkage disequilibrium (r2>0.5 or D’>0.5). Despite this, the pathway analysis was robust enough to highlight the associations of these two genes to smoking
