Both 5HTTLPR and rs25531 were genotyped using the assays described in Wray et al. [2009]. The assay for 5HTTLPR (which includes duplicate genotyping of each sample) was found to be more accurate than the original assays [Heils et al., 1996] which generate considerable bias towards S allele identification [e.g., Kaiser et al., 2002; Yonan et al., 2006]. Our data provided extensive opportunity for quality control checking: in total 6,607 DNA samples were genotyped (4,949 were the twins analyzed herein), of which 764 samples were duplicates (0.45% were discordant), 857 were from MZ twin pairs (0.16% were discordant), and some were from nuclear families [see Wray et al., 2009]. When only one individual from an MZ twin was genotyped, missing genotypes were imputed from the genotype of their co-twin: this recovered 264 and 280 genotypes for 5HTTLPR and rs25531, respectively. After exclusion of identifiable errors, genotyping call rates were 96.9% for 5HTTLPR and 95.4% for rs25531, somewhat lower than normal for our laboratory, but nonetheless much improved over the original assay for 5HTTLPR. In total, 1,232 individuals had been genotyped for
