Abstract
Background
Increasing diversity in genomic studies is critical for defining the genetic architecture of LOAD by improving power to identify variants more prevalent in or specific to a given ancestry. In this study, we constructed and analyzed a multi‐ancestry collection of GWAS datasets in the Alzheimer’s Disease Genetics Consortium (ADGC) to identify novel LOAD susceptibility loci and characterize shared and unique features of LOAD genomic risk profiles between ancestry groups.
Method
The ADGC multi‐ancestry dataset collection includes GWAS genotype and phenotype data on 38,774 non‐Hispanic White (NHW), 7,454 African American (AA), 11,436 Hispanic (HI), and 3,277 East Asian (EAS) subjects, all imputed to the NHLBI TOPMed v5 haplotype reference panel. We performed a two‐stage analysis: (1) single‐variant association analyses using score‐based logistic regression for case‐control and cohort studies and generalized linear mix‐model for family‐based datasets with covariate adjustment for onset/exam age, sex, principal components for population substructure, and APOE ε2/ε3/ε4 genotype, followed by within‐ancestry fixed‐effects meta‐analysis using METAL; and (2) cross‐ancestry meta‐analysis of within‐ancestry summary statistics using the random‐effects model (RE2) in METASOFT.
Result
In addition to APOE region associations, we identified five loci with cross‐ancestry genome‐wide significant associations (P≤5×10−8) including chromosomes 2q14 (BIN1; P = 2.4×10−19), 7q22 (NYAP1; P = 4.9×10−8), 11p2 (LRRC4C; P = 4.6×10−8), 11q12 (MS4A6A; P = 2.6×10−8), and 11q14 (PICALM; P = 2.3×10−10). 21 loci reached suggestive significance (P<10−5), including 10 associations driven by AA ancestry in or near ABCA7, APP, TAF1B, and CEP44); five by NHW (CR1, CD2AP, SORL1, ECHDC3, and ABCA1); four by HI (TREM2, IQCK, DIRC3, and FAR1); and one by EAS (VPS41), some with highly heterogeneous cross‐ancestry associations. Follow‐up analyses including cross‐ancestry fine‐mapping, gene‐based analyses, eQTL (expression) analyses, and functional analyses are in progress.
Conclusion
Cross‐ancestry GWAS meta‐analyses identified a novel LOAD susceptibility locus, LRRC4C, as well as a number of suggestive loci driven by individual ancestry groups. LRRC4C (Leucine‐Rich Repeat Containing 4C; MIM:608817) encodes NGL1, a ligand of Netrin‐G1 in the netrin family of axon guidance molecules, and has been shown to regulate development and function of thalamocortical axons. Multi‐ancestry studies with even larger sample sizes will provide even more powerful for further elucidating the genomic underpinnings of LOAD.