a connection between two other CHRN genes (CHRNA4 and CHRNA10) emerged from our secondary analysis of all SNPs that had been genotyped. The CHRNA4 gene (rs3787140) reached a similar level of significance as CHRNB3 for the top individual SNP empirical p-value (0.011) and gene-based empirical p-value (0.020), but was not significant experiment-wide (p=0.19). The most significant result was for SNP rs2231532 (p=0.002) in CHRNA10 (p=0.007), which reached experiment-wide significance (p=0.048) after correcting for all 226 SNPs. This is a novel association finding that was not observed in the analysis of case-control status in the overlapping NICSNP sample (uncorrected p=0.18) (Saccone et al., in press).
