Details

Autor(en) / Beteiligte

• Naj, Adam C.
• Sha, Jin
• Zhao, Yi
• Leonenko, Ganna
• Jian, Xueqiu
• Grenier‐Boley, Benjamin
• Dalmasso, Maria Carolina
• van der Lee, Sven J.
• Sims, Rebecca
• Chouraki, Vincent
• Bis, Josh C.
• Kuzma, Amanda B.
• Kunkle, Brian W.
• Karamujić‐Čomić, Hata
• Pitsillides, Achilleas N.
• Xia, Rui
• Fulton‐Howard, Brian
• Holmans, Peter
• Dupuis, Josée
• Wang, Li‐San
• Farrer, Lindsay A.
• van Duijn, Cornelia M.
• Haines, Jonathan L.
• Destefano, Anita L.
• Pericak‐Vance, Margaret A.
• Ramirez, Alfredo
• Amouyel, Philippe
• Lambert, Jean‐Charles
• Seshadri, Sudha
• Williams, Julie
• Schellenberg, Gerard D.

Titel

Genome‐wide meta‐analysis of late‐onset Alzheimer’s disease using rare variant imputation in 65,602 subjects identifies risk loci with roles in memory, neurodevelopment, and cardiometabolic traits: The international genomics of Alzheimer’s project (IGAP)

Ist Teil von

• Alzheimer's & dementia, 2020-12, Vol.16, p.n/a

Erscheinungsjahr

2020

Link zum Volltext

Quelle

Wiley Online Library - AutoHoldings Journals

Beschreibungen/Notizen

• Background Recent meta‐analyses of genome‐wide association studies (GWAS) have identified ∼30 susceptibility LOAD loci in addition to APOE, however the majority are common variants (minor allele frequency (MAF)>0.02). We used the dense, high‐resolution Haplotype Reference Consortium (HRC) r1.1 reference panel (64,976 haplotypes/39,235,157 SNPs), which allows imputation of rare variants (MAF>0.0008), to impute 44 GWAS datasets of the IGAP consortia to identify novel rare variant, gene, and pathway associations. Method We imputed 25,192 cases and 40,410 controls to the HRC r1.1 panel using Minimac3 on the Michigan Imputation Server. Converting imputed genotype probabilities to minor allele dosage, we ran logistic regression using SNPTEST on individual variants with MAF > 0.01 (and using generalized linear mixed models in R with family‐based datasets), and performed a fixed‐effects, inverse‐variance‐weighted meta‐analysis using METAL. Variants with MAF ≤ 0.01 were meta‐analyzed using score‐based tests via SeqMeta/R. Both analyses adjusted for age‐at‐onset(cases)/age‐at‐exam(controls), sex, and principal components for population substructure. Gene‐based associations were done with SKAT‐O and burden testing, while pathway associations were examined using VEGAS2. Result Discovery analyses of ∼39.2M genotyped or imputed SNVs confirmed single variant associations in 26 of 30 known IGAP LOAD loci at suggestive levels of significance (P < 10−5), with 12 known loci attaining genome‐wide statistical significance (P < 5 × 10−8) (Figure 1). Newly observed associations included common variants (MAF>0.01) in or near homologs of known AD loci, EPHA5 (rs17086136, OR[95% CI] = 1.23 [1.13,1.33], P = 6.36 × 10−7) and ADAM28 (rs10096379, OR[95% CI] = 0.86 [0.81,0.92], P = 3.02 × 10−6); in/near neuronal development genes including DAB1 (neuronal migration; 1:57700874:T:G, OR[95% CI] = 0.71 [0.62, 0.81], P = 6.94 × 10−7) and DCC (axon guidance; rs2054289, OR[95% CI] = 0.71 [0.62, 0.83], P = 6.69 × 10−6); and in/near genes involved in cardiometabolic traits including THADA (type 2 diabetes; rs77101426, OR[95% CI] = 0.89 [0.85, 0.95], P = 2.37 × 10−6). Several known AD loci demonstrated novel rare variant associations with genome‐wide significance, including CR1, PICALM, and the MS4A region (Figure 2), and novel rare variant associations were observed in or near genes involved in memory and cognitive function, including HS3ST4 and DBX1. Independent replication in external datasets including the UK Biobank is on‐going. Conclusion Several novel LOAD candidate loci, including those with prior associations with neurodevelopment and cardiometabolic traits, were identified using high‐quality imputation of rare and low‐frequency variants in IGAP.