Premium
Comparative trans‐ethnic meta‐analysis of whole exome sequencing variation for Alzheimer’s disease (AD) in 18,402 individuals of the Alzheimer’s Disease Sequencing Project (ADSP)
Author(s) -
Naj Adam C.,
Rajabli Farid,
HamiltonNelson Kara L.,
Kunkle Brian W.,
Jian Xueqiu,
Wang Yanbing,
Peloso Gina M.,
Lin Honghuang,
Sarnowski Chloé,
Pitsillides Achilleas N.,
Satizabal Claudia L.,
Bush William S.,
Thornton Timothy A.,
Wijsman Ellen M.,
Seshadri Sudha,
Haines Jonathan L.,
Dupuis Josée,
Byrd Goldie S.,
Mayeux Richard,
Destefano Anita L.,
Martin Eden R.,
PericakVance Margaret A.,
Fornage Myriam
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.041583
Subject(s) - exome , exome sequencing , genotype , genome wide association study , genetics , biology , genetic association , ethnic group , allele frequency , locus (genetics) , medicine , single nucleotide polymorphism , gene , mutation , sociology , anthropology
Abstract Background Using sequencing from multi‐ethnic AD studies, the ADSP aims to identify genomic variation contributing to elevated risk of, or protection from, AD. We examined coding region variation in the ADSP WES (whole exome sequencing) Release 2 dataset, comprising jointly‐called genotypes on 12,135 non‐Hispanic White (NHW), 4,108 African American (AA), and 2,159 Hispanic (HI) samples, to characterize ethnic differences and similarities in AD risk profiles across 40 known AD susceptibility loci. Method The ADSP WES Release 2 dataset includes genotype data on 8,789 cases and 9,613 controls collected in nine datasets using 10 target capture kits jointly called by the Genomic Center for Alzheimer’s Disease (GCAD). To test genetic associations, we first combined genotype data that had been QCed (quality‐controlled) by ‘subset’ of source dataset and capture kit, excluded low quality variants and samples, and then stratified by race/ethnicity. Within race/ethnicity, we performed association analyses with the package GENESIS, accounting for relatedness and population substructure, with covariate adjustment for QC subset. Result In preliminary analyses of 6,572,710 coding region variants, only APOE region associations attained genome‐wide significance ( P < 5 × 10 −8 ) across strata. Examining 20 AD loci with coding region associations in two prior exome analyses (Sims et al. 2017 and Bis et al. 2019) and/or a recent large GWAS of AD (Kunkle et al. 2019), different variants demonstrated nominal associations ( P < 0.05) in TREM2 across ethnicities (NHW, rs75932628, P < 10 −40 ; AA, rs2234255, P = 2.56 × 10 −3 ; HI, rs115953314, P = 1.30 × 10 −4 ). No other variant/gene was associated across all ethnicities, though ethnicity‐specific association patterns were observed at two genes among NHW and AA, but not HI: ABI3 (NHW, 16:81910580:A:G, P = 4.7 × 10 −4 ; AA, rs374229872, P = 5.61 × 10 −3 ) and PLCG2 (NHW, rs142527437, P = 2.45 × 10 −3 ; AA, rs35031462, P = 0.01). Analyses of NHW revealed strong associations at AC099552.4 (rs1043915, P = 0.0039) and PILRA (7:155196965:G:A, P = 1.0 × 10 −5 ), and among HI, a modest association was observed at LDB3 (rs76615432, P = 0.01). Analyses examining co‐localization of coding region signals across strata using ethnicity‐specific LD structure are on‐going and will be reported. Conclusion NHW and AA share associations in previously identified AD genes: TREM2 , ABI3 , and PLCG2 . Other genes demonstrated suggestive ethnicity‐specific associations, but work is on‐going to determine if these reflect true ethnic differences.