To evaluate the sensitivity of estimates of SNP-based heritability to modeling approaches, we assessed the data in multiple ways. The first assessment of data included all affected and unaffected individuals born in Sweden, of whom most, but not all, were of Swedish/European genetic ancestry; use all 405,105 high quality, genotyped SNPs for analysis. The sampling in this first assessment of data is consistent with our previously-published, family-based analyses and will be our primary analytical approach. The second assessment of data pruned SNPs according to linkage disequilibrium (LD) to obtain a smaller set of 184,296 largely independent SNPs. The third assessment of data limited the sample to individuals of European genetic ancestry. The fourth assessment of data removed all individuals for whom there is also a fifth degree or greater relative in the sample. The fifth assessment of data analyzed only pairs of affected and unaffected individuals, matched on two dimensions of genetic ancestry using the function pairmatch in the package optmatch in R (1-to-1 fullmatch) (Supplementary Materials). The sixth assessment of data was conducted as was done for the fifth,
