Single SNP association results (FBAT p-values) generated for the primary analyses in two independent “discovery datasets” consisting of families with multiple cases of late-onset AD (North European FAD and Caribbean-Hispanic FAD datasets). After adjustment for multiple testing with an FDR level of 0.1, the cutoff p-values for significant association are 0.024 in the North European family dataset and 0.003 in the Hispanic family dataset. MAF = minor allele frequency. Supplemental Table 2A also shows allele and genotype frequencies. Nominally significant p-values are in bold; * = association is significant after correction for multiple testing. The alleles putatively associated with AD are depicted only for SNPs generating nominal p-values of p ≤ 0.10.
Separate confirmatory analyses were performed in four independent “replication cohorts” including two case:control datasets (of North European origin or from an Israeli-Arab inbred population isolate), and two familial datasets (siblings) from the MIRAGE study (Caucasian Americans and the other of African-Americans). Corrections for multiple testing were not applied in these directed replication analyses. Boxes highlight identical alleles that are nominally associated with disease in at least two independent datasets. ND –not determined. NA – not available. The p-values in the case:control cohort are for allelic association. Alleles putatively associated with AD are depicted only for those SNPs generating nominal p-values of p < 0.10. Complete data for all SNPs is contained in Supplemental Table 3.