and rs10498633, the original publication named two genes, HLADRB5
and HLA-DRB1, and SLC24A4 and RIN3, respectively. For each
of these SNPs, we included both named genes. In addition to GWAS
SNPs,we included genes that contain rare variants that affect risk for
AD, and APP, PSEN1, and PSEN2, which contain functional variants
that cause early-onset AD and possibly harbor additional variants
that affect risk for late-onset AD (Table 1). Finally, we counted the
number of minor alleles of known GWAS SNPs for each individual
and included the genotype counts in covariate files to be used when
we wanted to control for known GWAS SNPs. So an individual could
have a count of 0 (indicating the individual is homozygous for the
major allele), 1 (indicating the individual is heterozygous for the
minor allele), or 2 (indicating the individual is homozygous for the
minor allele).