Assay noise also fails to provide a convincing alternative explanation for the data reported herein. When we examined the overlap between the clustered positive SNPs and the 10% of the SNPs for which the correlations between observed and expected values in validating studies were poorest, we found about as many (15) as we would have expected to find by chance (13). The t tests used for assignment of primary nominal significance also correct for assay variability. There is thus no indication that assay noise provides the sole basis for the addiction-associated allelic variants identified herein.
