FunGen-xQTL Resources

Alzheimer's Disease GWAS Summary Data (Wightman)

SNP-level association summary statistics for Alzheimer’s disease (AD) from Wightman et al. 2021 in Nature Genetics. This page covers three meta-analyzed sets of summary statistics: all individuals, all individuals excluding 23andMe, and all individuals excluding 23andMe and UKBB.
Cohort
AD GWAS
Modality
GWASGWAS
Release status
Stagednot yet verified
Lead analysts
Oluwatosin OlayinkaSelect a name to email them

Access

Controlled access. Request data access on the AD Knowledge Portal; you need a Synapse account. See How to get the data.

WhereIDAccessWhat you get
Synapsesyn69670625ControlledFiles described below

Other Synapse IDs on this page syn69670626 syn69670630 syn69696846 syn69865816 syn69865824 syn70095142 syn70095143

shell
pip install synapseclient
synapse login
synapse get -r syn69670625

Contact

Oluwatosin Olayinka

Source files

The summary statistics are available from the NIAGADS FTP site. The original data (GRCh37, 23andMe individuals only) and a liftover to hg38 are held on the Columbia University (CU) cluster.

Path to SuSiE RSS Fine-mapping Objects

AD GWAS–xQTL Integration

AD GWAS–xQTL colocalization results (ColocBoost): syn69865816

AD GWAS–xQTL colocalization models: syn69670630

AD colocalization models (all GWAS studies): syn69670626

Integrated AD Gene-Level Summary

AD risk genes prioritized by xQTL + GVC: syn70095142

AD risk genes prioritized by xQTL + TWAS + GVC: syn70095143

Download source

The download source is not public.

File Schema

  • chromosome: chromosome ID
  • position: hg38 position (converted from hg19 with liftOver)
  • variant: variant ID in the form chr${chromosome}_${position}_${ref}_${alt}
  • alt: alternative effect allele
  • ref: reference allele
  • z - z-score of the SNP effect size
  • pvalue: p-value of beta
  • N:
  1. Summary statistics preprocessing: https://github.com/floutt/brain-xqtl-analysis/blob/main/analysis/Zhang_BU/susie_rss/summary_stat_qc_all.nb.html

Cohorts included in this study

  • 1,126,563 individuals = 90,338 (46,613 proxy) cases + 1,036,225 (318,246 proxy) controls
  • 77,779 cases + 554,893 controls from Jansen et al.
  • 12,559 cases + 481,332 controls not from Jansen et al: Finngen, GRACE, HUNT, BioVU, 23andme, Gothenburg H70 Birth Cohort Studies and Clinical AD from Sweden (Gothenburg), ANMerge.
  • Population: US and Europe (UK, Norway, Sweden, Iceland, Finland, Spain)

Supplementary Table 1: A list of the datasets included in the meta-analysis. The UKB data was generated with a continous phenotype so the case-control values are estimates where the number of individuals with phenotype values <1 are controls and >=1 are cases.

DatasetPopulationCasesControlsImputation panel
Included in Jansen et al. (2019) | | | |
IGAP*US & Europe21982419441000G phase 1
UKB (Proxy)UK46613318246HRC, 1000G, UK10k
DemGeneNorway16386059HRC
TwinGeneSweden2246321HRC
STSASweden320750HRC
deCODE (partially included)Iceland7002181573deCODE
Total-77779554893-
Not included in Jansen et al. (2019)
Finngen*Finland179872206SISu v3
GR@CESpain41203289HRC
HUNTNorway11567157HRC+HUNT
BioVUUS60036059HRC
23andMeUS38073598391000 Genomes Phase 3, UK10K, HRC
Gothenburg H70 Birth Cohort Studies and Clinical AD from Sweden (Gothenburg)Sweden7122523HRC
ANMergeEUR366259HRC
Total-12559481332-
Grand total-90,3381,036,225
*=publically available