Based on visual inspection of the regional plots (LocusZoom) of all genomic loci identified in SNP-based meta-analyses of the 13 neuroticism phenotypes, we decided to further examine the plausibility of 6 of these loci, as they appeared to be driven by only one or two SNPs (Supplementary Data 3, 4). First, we compared the allele frequencies of the lead SNPs in these loci that we observed in the UKB data to the allele frequencies reported by 1000 genomes and TOPMED, as a large difference might indicate genotyping errors (Supplementary Data 3). Second, we compared the P values from GWA analyses in both samples to see whether the signal was driven by a subset of the data. Finally, differential LD across datasets might be another indicator of genotyping errors. Therefore, we selected all SNPs in LD (r2 > 0.6) with the lead SNPs in the 1000 Genomes data, and compared the LD in the 1000 Genomes data with LD (for the same combinations of SNPs) in the UKB data (Supplementary Data 4). None of these checks suggested genotyping errors of these
