Association tests were performed for individual SNPs in PLINK [26] using the regression option (additive model) and including sex and age as covariates. The statistical power to detect a genetic effect size of 1% (for MAF of 0.25) was 90.8% [27]. Because a large number of measures and SNPs were investigated, a Bonferroni correction was applied that resulted in a new significance criterion of .0002. This was based on correcting for five independent factors; that is, the five major personality dimensions (anxiety and depression are aligned with the neuroticism factor) multiplied by the number of independent SNPs. Nyholt’s [28] Single Nucleotide Polymorphism Spectral Decomposition program was used to estimate SNP independence within genes, resulting in a correction for 46 independent SNPs.
