For the analyses in this study, we excluded all individuals of non-Caucasian ancestry (based on genetic principal components). Therefore, principal components from the 1000 Genomes reference populations36 were projected onto the called genotypes available in UK Biobank. Those participants whose projected principal component score was closest to the average score of the European 1000 Genomes (based on the Mahalanobis distance) were identified as European. We excluded European subjects with a Mahalanobis distance > 6 s.d. Additionally, subjects were filtered out based on relatedness, discordant sex, sex aneuploidy, and withdrawn consent. Finally, individuals were excluded from analyses if responses to more than 3 (of 12) of the neuroticism items were missing. The effective sample in our meta-analysis consisted of 380,506 individuals (205,556 females and 174,950 males; see Supplementary Table 5 for item-specific sample sizes). At the time of test completion, participants’ age ranged between 40 and 73 (M = 56.91; s.d. = 7.93).
