GWAS Catalog Inclusion Criteria
Scope
The GWAS Catalog indexes genome-wide association studies (GWAS) published in peer-reviewed journals and associated genome-wide summary statistics.
Association data indexed in the GWAS Catalog is extracted from scientific literature indexed in PubMed.
A weekly literature search identifies publications that meet the GWAS Catalog eligibility criteria. Literature search and triage are performed using the machine learning-assisted triage system LitSuggest, developed at NCBI.
The GWAS Catalog curates studies, samples, traits and top associations from eligible publications. Additional annotation, such as genomic mapping, is obtained from external resources. The GWAS Catalog does not generate primary association data.
Key inclusion rules
Studies are eligible if they:
-
Analyse the human genome.
-
Include a primary GWAS using array- or sequencing-based genotyping.
-
Analyse at least 100,000 pre-quality-control variants.
-
Perform genome-wide analyses without restricting variants to candidate genes.
-
May reuse previously published GWAS datasets as part of a new GWAS analysis.
-
May use large targeted arrays (for example Metabochip, Immunochip or Exome arrays), which are flagged in the genotyping technology field.
Studies are excluded if they:
-
Are not published in English.
-
Analyse non-human species.
-
Only investigate candidate genes.
-
Use custom gene-based arrays without a clearly described genome-wide backbone.
-
Analyse somatic variation (for example tumour samples).
-
Do not contain new GWAS analyses but only report analyses of publicly available summary statistics.
Top Association Inclusion Criteria
Which SNP-trait associations are included?
The GWAS Catalog extracts the most significant SNP from each independent locus reported in an eligible GWAS.
Significance threshold
Top SNP-trait associations are included only if the reported association has a p-value less than 1 x 10⁻⁵.
This threshold is used for inclusion of top associations in the GWAS Catalog.
Combined discovery and replication analyses
If a study reports a combined discovery and replication analysis, the GWAS Catalog records the combined effect estimate and combined p-value.
When no combined analysis is available
If no combined p-value is reported, the GWAS Catalog records the association from the largest sample size provided that:
-
The discovery stage has p < 1 x 10⁻⁵.
-
The replication stage has p < 1 x 10⁻⁵.
Discovery-only studies
If no replication cohort is included, significant discovery-stage associations are indexed.
Previously known loci
Previously reported loci are included if they:
-
Are statistically significant in the reported GWAS.
-
Are reported by the authors even if they were not selected for replication.
Summary Statistics
The GWAS Catalog accepts full genome-wide summary statistics submitted directly by authors (see submission instructions) or obtained by curators from external sources.
Summary statistics may accompany:
-
Peer-reviewed publications.
-
Preprints.
-
Unpublished studies submitted by authors.
Pre-publication submissions are displayed separately and are clearly labelled as not curator-reviewed. For example:
After journal publication, these studies are fully curated, annotated, extended to include top associations and incorporated into the main GWAS Catalog.
Eligibility criteria for summary statistics are described on the Summary Statistics methods page.
Sequencing-based GWAS
The GWAS Catalog includes sequencing-based association studies (seqGWAS). A review of sequencing-based association studies is available here:
Sequencing-based association study review
Genome-wide summary statistics are accepted for eligible sequencing studies, including aggregate analyses such as gene-burden tests. Summary statistics may be submitted following the submission instructions.
Top associations from aggregate analyses are not currently indexed for search.
Targeted Arrays
Large targeted arrays, including Metabochip, Immunochip and Exome arrays, are included when they assay more than 100,000 variants.
Summary statistics from targeted and exome array studies are accepted provided they contain more than 100,000 variants. See the submission instructions.
These studies are flagged as targeted-array studies within the GWAS Catalog.
Machine-readable Summary
| Policy | Rule |
|---|---|
| Human studies | Required |
| Minimum variants analysed | >=100,000 pre-QC variants |
| Candidate gene studies | Excluded |
| Genome-wide gene/aggregate summary statistics | Included |
| Genome-wide gene/aggregate top associations | Not currently included |
| Sequencing-based GWAS | Included |
| Large targeted arrays | Included and flagged |
| Top association inclusion threshold | p < 1 x 10⁻⁵ |
| SNP extracted per independent locus | Most significant SNP |
| Summary statistics | Accepted from published studies and unpublished author submissions |
Pilot Projects
The GWAS Catalog inclusion criteria are reviewed regularly, and pilot projects are used to evaluate potential extensions to the Catalog scope.
Current pilot work is described on the Pilot Projects page.
For publications by the GWAS Catalog team and additional background reading, see the Related Resources page.
last updated: 07/08/2026