Multi-ancestry meta-analysis and fine-mapping in Alzheimer's disease

Abstract

Genome-wide association studies (GWAS) of Alzheimer's disease are predominantly carried out in European ancestry individuals despite the known variation in genetic architecture and disease prevalence across global populations. We leveraged published GWAS summary statistics from European, East Asian, and African American populations, and an additional GWAS from a Caribbean Hispanic population using previously reported genotype data to perform the largest multi-ancestry GWAS meta-analysis of Alzheimer's disease and related dementias to date. This method allowed us to identify two independent novel disease-associated loci on chromosome 3. We also leveraged diverse haplotype structures to fine-map nine loci with a posterior probability >0.8 and globally assessed the heterogeneity of known risk factors across populations. Additionally, we compared the generalizability of multi-ancestry- and single-ancestry-derived polygenic risk scores in a three-way admixed Colombian population. Our findings highlight the importance of multi-ancestry representation in uncovering and understanding putative factors that contribute to risk of Alzheimer's disease and related dementias.

Fig. 1. Study Design.

Outline of multi-ancestry meta-analysis procedure and downstream analysis.

Fig. 2. Summary of multi-ancestry meta-analysis.

a The Manhattan plot for the random effects meta-analysis P values, truncated at −log10(P) < 50. An orange dot indicates that the lead SNP at a locus reached a P value < 5 × 10^-8, while a red dot indicates a P value < 5 × 10⁻⁹. b, c The corresponding local association plots for the two loci of interest and forest plots summarizing the effect estimates per ancestry group for lead SNPs at the two loci of interest. Lead SNPs from both novel loci were absent in the East Asian dataset by Shigemizu et al. used for discovery. Abbreviations - MR MR-MEGA, RE Random effects, FE Fixed effect, EUR European, EUR (FIN) Finnish European, AFR African American, CAR HISP Caribbean Hispanic.

Fig. 3. Graphical summary of heterogeneity at ADD genetic risk loci.

Lead SNPs were derived from MR-MEGA using maximal LD blocks, apart from APOE rs429358 and rs7412. Both APOE SNPs were absent in summary statistics from the most recent European-ancestry ADD GWAS. Aggregate effects were estimated using a random effects model since MR-MEGA assumes that effects differ across populations. Allelic effect heterogeneity that is attributable to genetic ancestry was estimated using Cochran’s Q statistics for ancestral and residual heterogeneity from the meta-regression (“Online Methods”).

Fig. 4. Graphical summary of genetic risk scores.

These genetic risk scores were derived from multi-ancestry and ancestry-specific risk estimates, then applied to an admixed Colombian cohort to evaluate significance and predictive power. The European-based (EUR) PRS was derived from a fixed effect meta-analysis of the summary statistics from Bellenguez et al. and FinnGen used in the meta-GWAS. a, b The maximal AUCs for each genetic risk score with color coding to delineate the source of the risk estimates for scores excluding and then including APOE-e4 variants. P value thresholds that correspond to the maximal AUCs are shown in Table S6.