Skip to main content

Frequently Asked Questions

Answers to common questions about the GWAS Catalog.

A. General​

B. Searching the GWAS Catalog​

B(i). Searching by publication​

B(ii). Searching by variant or gene​

B(iii). Searching by trait​

B(iv). Searching by study / study design​

C. Exploring specific pages​

D. Diagram​

E. Genomic mappings​

F. Population descriptors and sample metadata​

G. Programmatic access​

H. Summary statistics​

I. Submitting summary statistics​

A. General​

How can I access GWAS Catalog data?​

The GWAS Catalog contains data from published and unpublished human genome-wide association studies (GWAS). You can access GWAS Catalog data in the following ways:

  • Web search interface: Search by publication, study, trait, variant, or gene.

  • Specific entity pages: Each publication, study, trait, variant, or gene has its own page. Use the search bar to find an entity and then explore its page.

  • Downloadable spreadsheets: Full association and study data are on the file downloads page. You can also download associations directly from any entity page via the “Download Associations” button.

  • Full genome-wide summary statistics: Download from our FTP site or from the summary statistics page.

  • Programmatic access: The REST API documentation is here: /programmatic-access/rest-api.

  • Visualisation: The GWAS Catalog diagram provides a graphical view of the data: https://www.ebi.ac.uk/gwas/diagram.

Related search terms: data download, access, search interface, pages, programmatic access, visualization, API, FTP.

Who maintains and funds the NHGRI-EBI GWAS Catalog?​

From September 2010 onward, the GWAS Catalog is jointly developed and delivered by EMBL-EBI and NHGRI.

  • Founded by NHGRI.

  • Infrastructure moved to EMBL-EBI in March 2015 to support ontology-driven search, automated quality control, and improved curation tools.

  • The NHGRI site is no longer available

Related search terms: maintainers, funding, EMBL-EBI, NHGRI, project history, infrastructure.

Where does the GWAS Catalog’s data come from?​

The GWAS Catalog sources data using:

  • A weekly PubMed literature search using LitSuggest to identify publications that meet our eligibility criteria.

  • Manual curation of data — studies, samples, traits, and significant associations - from PubMed-indexed publications.

  • External data integration — ontology annotation, genomic mapping, and other metadata from external sources.

  • Author submission of unpublished studies (since 2020) — pre-prints, in-press, or standalone GWAS with full summary statistics and metadata.

  • For more detail, see our Methods section.

Related search terms: literature search, PubMed, curation, LitSuggest, unpublished GWAS, submission.

How can I learn more about using the GWAS Catalog and stay informed?​

Training is available through:

Related search terms: training, mailing list, social media, help, resources.

How should I cite the GWAS Catalog?​

When citing the GWAS Catalog in a publication, use the most recent official GWAS Catalog publication and include the date you accessed the data. The latest publication is listed on our About page.

If you are citing specific studies or summary statistics from the GWAS Catalog, include:

  • The study accession ID (e.g., GCST000123).

  • The GWAS Catalog name.

  • The date downloaded.

  • The original publication if available.

Full citation guidance is also available on the About page.

Related search terms: cite, reference, publication, GWAS Catalog paper, Nucleic Acids Research, GCST accession

What are the terms and conditions for accessing the GWAS Catalog data and code?​

Summary statistics: GWAS Catalog summary statistics are released under the Creative Commons CC0 1.0 Public Domain Dedication, unless otherwise specified on the individual study page and summary statistics download page.

Other GWAS Catalog data: Other data in the GWAS Catalog can be used under the EBI Services Terms of Use.

Code: Our code is available under the Apache License 2.0.

Related search terms: terms of use, CC0, license, licensing, Apache 2.0, public domain, data sharing, usage restrictions

How often is the GWAS Catalog updated?​

Data release schedule: The GWAS Catalog adds new curated data approximately every two weeks. Each data release includes updates to the website and all downloadable spreadsheets available on the file downloads page. The date of the most recent release is shown at the bottom of the GWAS Catalog home page.

Summary statistics availability: Summary statistics files are made available as soon as possible, even before the related study appears in a scheduled data release. If a manuscript states that summary statistics are available from the GWAS Catalog, but you cannot find them in the list of studies with summary statistics files or on the FTP site, please contact gwas-info@ebi.ac.uk for further assistance.

Diagram The Diagram is updated quarterly. The data refresh date is shown in the Diagram page header.

Related search terms: update schedule, release frequency, data refresh, biweekly updates, summary statistics availability, missing files, diagram

Where can I find the GWAS Catalog code and software?​

The GWAS Catalog develops and maintains a range of open source software, data standards, and tools, which are available from our GitHub organisation: https://github.com/gwas-catalog

Related search terms: code, GitHub, software, open source, standards, repository

Does the GWAS Catalog include data that is not linked to a PubMed-indexed publication?​

Yes. Since 2020, the GWAS Catalog has accepted submissions of pre-published GWAS. This includes:

  • Preprints.

  • Articles in press.

  • Standalone datasets not yet published in a journal.

Pre-published data is made publicly available as submitted by authors. It is not manually reviewed by curators until the study is published in a journal.

Where to find pre-published data:

Once a study is published, it is curated, annotated, and extended with top associations before being fully incorporated into the main database.

Related search terms: pre-published GWAS, preprint, in press, standalone dataset, submission, summary statistics, curation status

B. Searching the GWAS Catalog​

How do I search the GWAS Catalog?​

Enter your query (e.g. “breast carcinoma”) into the search box and press return or click the search icon.

The search will return results across multiple entity types:

  • Publications (marked with “P”)

  • Variants (marked with “V”)

  • Traits (marked with “T”)

  • Genes (marked with “G”)

  • Studies (marked with “S”)

  • Regions (marked with “R”)

You can refine the results using the “Refine search results” panel on the left to filter by entity type.

Related search terms: search, query, publication search, trait search, variant search, filter, refine results

B(i). Searching by publication​

How do I search by publication?​

You can search for a publication using:

  • The PubMed ID.

  • Any author name.

  • Words from the publication title.

All authors listed in a paper are indexed in the Catalog. Searching by author will return every publication that includes that name, not just first-author papers. This may produce a large number of results.

For the most precise results, use the PubMed ID.

Related search terms: publication search, PubMed ID, author search, title search, publication results

Why do author searches also return traits?​

The search function matches your text across all fields in the Catalog.

For example, a search for “Parkinson” will return:

  • Publications with an author named Parkinson.

  • Publications with “Parkinson” in the title.

  • Traits related to Parkinson’s disease.

If you only want a specific publication, use its PubMed ID for the most accurate result.

Related search terms: author search, trait overlap, publication search, text match, PubMed ID

What is a study in the GWAS Catalog? How is it different from a publication?​

In the GWAS Catalog, a publication is an article published in a scientific journal. Each publication is tracked in the GWAS Catalog using its unique PubMed ID.

A study refers to a specific genome-wide association analysis within a publication. One publication may include several studies if it reports distinct analyses (for example, different traits, sample cohorts, or statistical designs).

Each study in the Catalog is assigned a stable accession number beginning with "GCST" (e.g., GCST000123). This accession uniquely identifies the analysis regardless of the publication it came from.

A study is the primary unit of curation in the GWAS Catalog.

For more details on how we handle publications containing multiple analyses, see our Curation methods section.

Related search terms: publication, study, what is a study, study definition, PubMed ID, GCST accession, multiple analyses, difference

Why is a newly published GWAS paper not yet in the GWAS Catalog?​

Every publication added to the GWAS Catalog goes through detailed manual curation. This process takes time, so there is usually a delay between when a paper is first indexed in PubMed and when it appears in the Catalog.

The curation team works as quickly as possible while maintaining high accuracy standards. In most cases, new publications take several weeks to a few months before being included.

If a paper of interest is more than two months old and not visible in the Catalog, please contact gwas-info@ebi.ac.uk to confirm it has been identified and is in the curation queue. Publications of particular scientific or community interest may be prioritised.

Related search terms: delay, missing study, new publication, curation backlog, PubMed indexing, processing time

B(ii). Searching by variant or gene​

How do I search by variant?​

You can search for a variant (single nucleotide polymorphism, SNP) in several ways:

  • By rsID (e.g., rs12345).

  • By genomic region (e.g., 6:16000000-25000000 or 2q37.1).

  • By a gene that is mapped to the variant.

Searching by gene or region may return many results, so if you are looking for one specific variant, the rsID is the most precise option.

Variants are mapped to genes using the GWAS Catalog’s genomic annotation pipeline, which includes Ensembl-based mapping of nearby or overlapping genes.

Related search terms: variant search, SNP, rsID, genomic region, gene mapping, Ensembl annotation

How do I search by gene?​

Enter a gene symbol (e.g., STAT4) into the main search bar. The search will return:

  • The gene itself.

  • Variants annotated with that gene by the Catalog’s mapping pipeline.

  • Publications that include the gene name in the title.

The “Gene” page lists all associations mapped to that gene and provides related gene-specific information.

Note: the mapped gene may not always match the gene reported by authors. Different criteria can be used for reporting, so author-reported genes may differ.

Author-reported genes were curated until 2021. They can still be found in the full data download. To view them:

  • Open the file in Excel.

  • Filter the REPORTED GENE(S) column for your gene of interest.

You can also use the REST API to retrieve associations by gene. See the API guide for details about mapped-gene filters.

Related search terms: gene search, gene page, reported gene, mapped gene, REPORTED GENE(S), REST API, curation cutoff 2021

Can I use the GWAS Catalog API to query by genomic region?​

Yes. REST API v2 supports filters by genomic location. See the API reference for supported endpoints and query syntax.

On the website, you can search by genomic region using either:

  • Chromosome coordinates in the format chromosome:start-end (e.g., 6:16000000-25000000).

  • Cytogenetic band notation (e.g., 2q37.1).

Both methods return a list of genes and variants located within the specified region.

Related search terms: genomic region API, REST API location filter, chromosome position query, region search, chromosomal coordinates, cytogenetic band, SNP location, genomic interval

What are "top associations"?​

Top associations are the curated set of associations displayed in the Catalog’s Associations tables. They are manually extracted from published articles and filtered using the GWAS Catalog curation process.

To be included as a top association, a result must:

  • Be significant at p < 1 × 10⁻⁵ in all stages of the analysis.

  • Be described as independent by the authors, or represent the peak association within a 100 kb region.

Top associations differ from full summary statistics. Summary statistics include all variants tested in a discovery GWAS, regardless of independence or significance, while top associations provide a focused list of the most relevant results.

Related search terms: top associations, significant SNPs, independent signals, peak SNP, summary statistics, curation criteria

How can I find mouse orthologs for GWAS Catalog genes in IMPC?​

Each gene in the GWAS Catalog that has a known mouse ortholog in the International Mouse Phenotyping Consortium (IMPC) is linked via a button on the gene page.

Orthology predictions come from IMPC’s reference database, which is updated weekly to include the latest HCOP ortholog relationships and data from MGI.

If no ortholog has been established for a gene, the IMPC button will not appear.

For details on ortholog mapping, see the IMPC documentation and the publication Cacheiro et al., Nature Communications (2020).

Related search terms: mouse ortholog, IMPC, HCOP, MGI, gene mapping, orthology

How can I separate odds ratios (OR) from beta values in the associations download?​

In the full Catalog download, odds ratios (OR) and beta values share the same column. They can be distinguished because beta values always include a unit and direction (e.g. “unit increase,” “cm decrease”), which is recorded in the 95% CI (TEXT) column.

If you need OR and beta values in separate columns, use the download option from the Associations table on the web interface. The exported file mirrors the table view, where OR and beta appear in different columns.

Related search terms: odds ratio, beta, association file, download, effect size, confidence interval

Is each association in the GWAS Catalog unique, and what is p-value annotation?​

Each association in the Catalog comes from a distinct analysis.

However, in older publications associations for the same variant may appear more than once within a study because:

  • Component groups of a meta-analysis may be represented individually, in addition to the overall meta-analysis results.

  • Where multiple populations (e.g. men/women, adults/children, ancestry groups) were analysed for the same trait, they were included in a single study

  • Where multiple related traits were analysed, they were sometimes included in a single study

In these cases, the associations were labelled with “p-value annotation”.

Since 2020, all GWAS on a distinct trait/population combination have been represented as separate studies in the Catalog (this represents the vast majority of data in the Catalog).

For confirmation, always refer back to the source publication.

Related search terms: association uniqueness, duplicate associations, meta-analysis, cohort reuse, repeated analysis

How does the GWAS Catalog represent extremely small p-values?​

Some publications report association p-values as 0. This usually happens when the analysis software cannot represent extremely small numbers.

When authors cannot provide the exact value, the GWAS Catalog records the threshold instead. For example, if the publication reports p < 1e-300, the Catalog records it as 1e-300, but the true p-value may be much smaller than the recorded threshold.

Related search terms: p-value, extremely small p-value, zero p-value, threshold, statistical precision, significance

What p-value threshold does the GWAS Catalog use for top SNP associations?​

The GWAS Catalog indexes, curates and includes top SNP-trait associations with p < 1 x 10⁻⁵ that meet our detailed curation criteria. This is the Catalog’s inclusion threshold; it is not intended to define genome-wide significance. See the Top Associations criteria here: Methods: Top Associations criteria

Does the GWAS Catalog require genome-wide significance (p < 5 x 10⁻⁸)?​

No.

The GWAS Catalog uses p < 1 x 10⁻⁵ as its inclusion threshold for top associations. This threshold determines which associations are curated and indexed. It is not intended to define genome-wide significance.

B(iii). Searching by trait​

How do I search by trait?​

Type the name of a disease, phenotype, measurement, or drug response into the search bar.

The search will return:

  • Traits that exactly match your term.

  • Synonyms of matching traits.

  • Child traits (e.g., searching “cancer” also returns cancer subtypes).

  • Publications where the title contains your search term.

If you cannot find your trait of interest, it may be stored under a synonym. For example, searching “general cognitive ability” returns “intelligence,” which is the standardised trait name in the Catalog.

The search bar also offers auto-suggestions, including synonyms, as you type.

Related search terms: trait search, phenotype, synonym, child term, ontology, auto-complete

When I search for a trait, do the results include all publications for that trait?​

No. A publication will only appear in search results if the title, author list, or PubMed ID contains your search term.

To see all studies linked to a trait, go to the trait’s dedicated page and view the Studies table. This table lists every study associated with that trait, even if the publication title does not mention it.

Related search terms: trait search, publication results, study table, PubMed ID, trait page

Why do other traits appear in my search results?​

Search results can include more than just exact matches to your term. They may also include:

  • Synonyms and child traits — for example, searching “thyroid disease” also returns “Hashimoto’s thyroiditis” and “Graves disease,” which are subtypes of thyroid disease.

  • Traits studied in combination with your trait — for example, searching “asthma” also returns “response to bronchodilator,” because a GWAS studied that response in individuals with asthma.

  • Publications — if the publication title contains your search term.

These connections come from the Experimental Factor Ontology (EFO), which defines synonyms, hierarchical relationships, and trait combinations. This allows the Catalog to support more comprehensive searches, even when studies describe traits in different ways.

Related search terms: trait search, child trait, ontology, synonyms, compound trait, EFO

How are multiple or compound traits represented in the Catalog?​

Some studies are mapped to more than one trait when the phenotype definition is complex. The Catalog records both standardised ontology traits and the reported trait taken directly from the publication.

  • Combined groups (“or”): If a study compares groups with different traits to a shared control group, the reported trait uses “or.” Example: “bipolar disorder or schizophrenia” → mapped to both “bipolar disorder” and “schizophrenia.”

  • Individuals with multiple traits (“and”): If a study analyses individuals with more than one trait, the reported trait uses “and.” Example: “bipolar disorder and binge eating” → mapped to both “bipolar disorder” and “binge eating.”

  • Biobank and code-based traits: Reported traits that include "UKB data field" or ICD codes may appear exactly as written in the publication, due to issues of scale. These cases may not follow our standard naming conventions. Users should consult the source (e.g., https://biobank.ndph.ox.ac.uk/ukb/index.cgi) to confirm the details.

Related search terms: compound trait, multiple traits, reported trait, ontology, UK Biobank, ICD codes

What is a background trait?​

A background trait is a characteristic shared by all participants in a study, but not directly tested in the association analysis.

For example:

In a study of “Allergic rhinitis in asthma,” the comparison is cases (individuals with allergic rhinitis) vs. controls (individuals without allergic rhinitis). All participants have asthma.

  • The main trait under investigation is allergic rhinitis.

  • The background trait is asthma.

The study cannot identify variants associated with asthma itself, but the fact that all participants have asthma provides important context for interpreting the allergic rhinitis associations.

The GWAS Catalog displays both main and background traits, clearly labelled, on study and association records. Background traits are also included in the most recent spreadsheet downloads (v1.0.3).

Related search terms: background trait, main trait, study context, phenotype definition

How are background traits represented in the Catalog?​

A background trait is a characteristic shared by all participants in a study, but not directly tested in the association analysis. Example: In a study of “Allergic rhinitis in asthma,” the main trait is allergic rhinitis, while asthma is the background trait.

Representation in the web interface:

  • On each Study page, background traits are shown in a separate field from the main trait.

  • Studies and Associations tables include columns for both main and background traits.

  • The Trait page, by default, shows only main traits. To also include studies where that trait is a background trait, tick the “Include background traits data” box above the data tables. This also updates the association plot.

In downloads:

  • Background traits are included in the most recent spreadsheet version (v1.0.3) under the columns MAPPED BACKGROUND TRAIT and MAPPED BACKGROUND TRAIT URI.

  • Earlier spreadsheet versions only include main traits.

Programmatic access:

  • The REST API currently only returns main traits.

Related search terms: background trait, main trait, study context, spreadsheet, API

Why did my search return no results even though I expected a match?​

The GWAS Catalog search looks for exact text string matches. Small differences in wording or spelling can affect results.

Examples:

  • Typing “beast cancer” instead of “breast cancer” will return nothing.

  • Searching “metabolic disorder” gives no matches, while “metabolic disease” returns many.

Tips if your search fails:

  • Use the auto-complete suggestions offered in the search bar.

  • Try alternative terms or synonyms.

  • Check the Experimental Factor Ontology (EFO) at http://www.ebi.ac.uk/efo to see if your trait is stored under a different name.

Related search terms: no results, spelling error, synonyms, auto-complete, EFO

What is the difference between a trait and a reported trait?​

A trait in the GWAS Catalog is a standardised term from the Experimental Factor Ontology (EFO). Traits represent the disease, phenotype, measurement, or drug response being investigated. Each trait has its own page in the Catalog, showing all relevant studies and associated variants. For more about how ontologies are used, see our ontology documentation.

A reported trait is taken directly from the authors’ description of the phenotype analysed in the study, and is provided in addition to the EFO trait. Reported traits are free text descriptions reflecting the study design and context, and may include additional details such as environmental interactions or compound phenotypes.

Together, the standardised trait terms support consistent searching and comparison across studies, while the reported traits preserve the exact wording and design used by study authors.

Related search terms: trait, reported trait, phenotype, EFO, ontology, study design, author description

B(iv). Searching by study​

Can I search for a study?​

To find a study:

  • Navigate to a publication, trait, or variant, then open the Studies table to see linked studies.

  • If you already know the accession number (e.g. GCST000123), you can enter it in the search bar on the homepage. This will return the study itself and its associated publication.

Related search terms: study search, GCST, accession number, study table

How can I search for targeted and exome array studies and associations?​

To find targeted or exome array studies:

In the search bar: Enter the genotyping technology term (e.g. “targeted genotyping array” or “exome genotyping array”). This will return publications that used that array.

In the user interface: Targeted and exome array studies are flagged with a small target icon in search results and in the Study accession column of the Studies table.

In downloads: The downloadable studies file includes a column called Genotyping technology (additional array information), which records whether a study used a targeted genotyping array, exome array, or similar technology.

Related search terms: targeted array, exome array, genotyping technology, icon, study accession, download file

How do I find the accession number for a study?​

Each study in the GWAS Catalog has a unique accession number beginning with “GCST” (for example, GCST000123).

You can find study accession numbers in several places:

  • At the top of each “Study” page.

  • In the “Studies” and “Associations” data tables on Publication, Trait, Variant, and Gene pages.

  • In the downloadable spreadsheets (associations, studies, and ancestry files) from version v1.0.2 onwards.

Note: Accession numbers are not included in the older v1.0 spreadsheets, which were only retained for backwards compatibility with the original NHGRI spreadsheet format.

Related search terms: accession number, GCST, study ID, identifier, download file, legacy format

How do I identify interaction studies in the GWAS Catalog?​

The GWAS Catalog includes SNP-by-environment interaction studies provided the SNPs meet genome-wide significance criteria. SNP-by-SNP studies were included up to 2023.

Identification in the UI and download files:

  • On the Studies page, tick the GxE checkbox under "Select study type:" to filter the table to only Gene-by-Environment interaction studies.

  • The GxE column in any studies table indicates whether a GxE interaction study design (tick or “x”). This column can also be exported via the file or enabled/hidden using Column visibility.

  • The downloadable studies file includes the GxE field so you can filter programmatically.

Identification via reported trait:

  • SNP-by-environment interactions include the environment in the description. As of July 2018, you’ll also see notation for test types: “(1df test)” or “(2df test)”, such as “Lung cancer × smoking interaction (1df test)”. Earlier studies may only say “(smoking interaction)”.

  • SNP-by-SNP interactions include “SNP × SNP interaction” in parentheses.

Alternate navigation:

  • On any “Trait” page, search “interaction” in the “Associations” or “Studies” tables.

  • In downloadable spreadsheets, filter the reported trait column for "interaction".

Related search terms: GxE checkbox, interaction studies, filter, SNP-by-environment, SNP-by-SNP, 1df, 2df, trait search

How do I identify targeted and exome array studies in the GWAS Catalog?​

Targeted and exome array studies are flagged in several ways:

In the user interface:

  • A small “target” icon appears in search results next to any publication that includes a targeted or exome array study.

  • The same icon is shown in the “Study accession” column of the Studies table on Publication, Trait, Variant, or Gene pages.

In downloads:

  • In the downloadable studies file, these studies include an additional column called Genotyping technology (additional array information).

  • This field records whether the study used a targeted genotyping array, exome array, or similar technology.

For details on available download fields, see the file downloads documentation.

Related search terms: targeted array, exome array, genotyping technology, icon, study accession, download file

Does the GWAS Catalog include CNV studies?​

We are working on expanding the scope of the GWAS Catalog to include genome-wide copy number variant (CNV) studies.

Researchers may submit CNV study summary statistics directly to the Catalog. Submitted files and metadata are made publicly available, even though CNV associations are not yet curated in the same way as SNP-based GWAS.

Related search terms: CNV, copy number variation, structural variation, submission, summary statistics, study scope

Are results from sequencing-based association analyses included in the GWAS Catalog?​

Yes — sequencing-based association analyses are now routinely curated, submitted and included in the GWAS Catalog, alongside array-based analyses.

We curate significant associations from sequencing-based studies using the same manual process as for array-based GWAS, ensuring consistent quality and annotation.

You can read more about the standards for reporting sequencing-based GWAS in McMahon et al., Sequencing-based genome-wide association studies reporting standards, Cell Genomics (2021).

You can identify sequencing-based studies by downloading the studies file from the file downloads page and filtering the Genotyping technology (additional array information) column for sequencing-based technologies (e.g., whole-genome sequencing, exome sequencing).

Related search terms: sequencing-based GWAS, whole-genome sequencing, WGS, exome sequencing, curated studies, reporting standards, genotyping technology filter

C. Exploring specific pages​

How do I download associations for a trait, publication, study, or variant?​

You can download association data from any Trait, Publication, Study, Variant, or Gene page in two ways:

Full data download: Click the Download Associations button to download a .tsv file containing all associations displayed on that page. This file uses the same format as the full GWAS Catalog spreadsheets from the file downloads page and includes study information for each association.

Customisable table download: The Studies and Associations tables show a simplified view of the data. Use the Export button in the top right corner to download a .csv version of the table. You can add or remove columns using Column visibility (only selected columns will appear in the exported file).

Related search terms: download associations, studies table, TSV, CSV, export, file downloads

How do I use the association plot on a Trait page?​

The association plot shows all associations for the selected trait as circles on a genome-wide view.

Circles: Each circle represents a variant–trait association.

Colour coding: Circles are coloured by broad trait category (see the legend in the top-left corner of the plot).

Interaction: Hover or click on a circle to view detailed information about that variant (e.g., p-value, mapped gene, study).

Download: Use the download button to save an image of the plot.

The association plot is powered by the LocusZoom plugin.

Related search terms: association plot, trait page, visualisation, LocusZoom, genome-wide view, download plot

How do I use the linkage disequilibrium (LD) plot on a Variant page?​

The LD plot shows the strength of linkage disequilibrium between the selected variant and nearby variants, using data from Ensembl.

Controls:

  • Choose a population and LD metric (r² or D′) from the dropdown menus.

  • Adjust the LD threshold to focus on stronger or weaker correlations.

Download: You can download the LD data as a .tsv file directly from the plot interface.

Programmatic access: LD data can also be retrieved via the Ensembl REST API: http://rest.ensembl.org/documentation/info/ld_pairwise_get. You can specify the variant ID, window size, population, and threshold. For multiple variants, make separate API calls.

Related search terms: LD plot, linkage disequilibrium, r2, D’, Ensembl, variant page, REST API, download LD data

How do I view, search, and download data from the Associations and Studies tables?​

The Associations and Studies tables appear on every Trait, Variant, Gene, Publication, and Study page. These tables are fully customisable:

  • Filter: Use the filter fields at the top of individual columns to narrow results to specific values.

  • Customise columns: Use the “Column visibility” (or “Add/Remove Columns”) menu to show or hide columns.

  • Sort: Click any column header displaying up-and-down arrows to sort by that column.

  • Download: Use the “Export data” button to download the current view as a CSV file. Your search, filters, and selected columns are preserved in the export.

Related search terms: studies table, associations table, filter, column visibility, sort, export csv, download

Does the GWAS Catalog provide Manhattan plots?​

No. The GWAS Catalog does not currently provide Manhattan plots.

We do provide LocusZoom plots for traits and LD plots for associations where available.

Related search terms: Manhattan plot, LocusZoom, LD plot, trait plots, association plots

D. Diagram​

How do I display SNP information for a given dot?​

Click on any trait (a coloured circle) on the diagram to open a pop-up. The pop-up shows:

  • all SNPs linked to that trait in that location

  • the p-value for each SNP–trait association

  • the study where the association was reported

  • the trait as curated in the GWAS Catalog

  • the mapped Experimental Factor Ontology (EFO) term

Each field in the pop-up is interactive. SNPs, traits, EFO terms and studies link to Catalog searches, and publications link to EuropePMC. Close the pop-up by clicking outside it, or by using the cross or "Close" button inside.

Related search terms: interactive diagram, diagram pop-out

How do I filter the diagram?​

You can filter the diagram by clicking trait categories in the sidebar on the left. At this time, filtering by individual traits is not supported. If you need more specific trait filtering, contact us at gwas-info@ebi.ac.uk.

Related search terms: diagram filter, diagram trait category, diagram sidebar filter

What are the trait categories shown on the diagram?​

The sidebar includes a colour legend with a count of dots in each category. Clicking a category highlights only those associations.

Some traits belong to multiple categories (e.g. Crohn’s disease is both a digestive system and immune system disease). Each dot is assigned a single colour, based on its most specific ancestor in EFO (ancestor that has itself the most number of ancestors).

Related search terms: diagram trait categories, diagram colour legend, diagram EFO ancestor, diagram disease categories

How do I display labels?​

Hover over chromosomes or coloured circles to see labels. Labels show the EFO term mapped to the SNP.

Related search terms: diagram labels

How should I cite the diagram?​

Please see the GWAS Catalog About page for citation guidance.

Related search terms: diagram citation

How do I download the diagram?​

Download options are available on the GWAS Catalog website at: Downloading the GWAS Diagram.

Related search terms: diagram download, image export, diagram formats

Can I download a filtered version of the diagram?​

Currently, the GWAS Catalog does not provide a way to directly download trait-filtered diagrams. The easiest option is to take a screenshot.

Related search terms: filtered diagram download, trait subset diagram, screenshot

E. Genomic mappings​

How is the genomic annotation for each SNP provided?​

The GWAS Catalog uses an Ensembl-based mapping pipeline to add genomic information for each SNP, including chromosome position, cytogenetic band, and mapped genes.

If a SNP falls inside a gene, that gene is reported. If the SNP is intergenic, the nearest upstream and downstream genes are reported.

Mapping information is refreshed with every Ensembl release (roughly every 2–3 months).

For more detail, you can use the GWAS Catalog REST API to retrieve the closest upstream and downstream genes reported from Ensembl and NCBI RefSeq.

Related search terms: SNP mapping, genomic annotation, Ensembl pipeline, gene mapping, intergenic SNP, upstream gene, downstream gene, genome build

Which genome build is used in the GWAS Catalog?​

Data in the GWAS Catalog is currently mapped to genome assembly GRCh38.p14 and dbSNP Build 156.

This information is updated on release and displayed at the bottom of the GWAS Catalog’s home page.

Related search terms: genome build, GRCh38, dbSNP build 156, reference genome, assembly version

How can I access GWAS Catalog data on alternative genome builds?​

The GWAS Catalog itself uses GRCh38, but you can remap SNPs to older builds such as GRCh37.

Options for remapping:

Related search terms: alternative genome builds, GRCh37, liftover, Ensembl REST API, assembly mapping, rsID conversion

Why is there no mapping for some associations in the GWAS Catalog?​

Here, mapping means genomic mapping of an association’s SNP to coordinates or genes, not mapping a trait to an ontology term. An association may be listed even when its SNP has no mapped genomic coordinates or mapped genes.

Possible reasons:

  • Typos or unusual identifiers: If a publication contains a typographical error or reports a non-standard SNP ID, the mapping pipeline may fail to find a match.

  • Coordinate-only variants: If a paper reports an association as chr:bp without an rsID, the Catalog records that value as the variant ID. The rsID-based mapping pipeline cannot process it, so mapping information is unavailable.

  • Retired SNPs: Some SNPs no longer exist in the latest genome build used by the Catalog. These SNP IDs are still displayed for completeness, but mapping data is not available.

Related search terms: missing association mapping, SNP genomic mapping, variant mapping, missing coordinates, coordinate-only variant, chr:bp, no rsID, unplaced SNP, dbSNP retired, unmapped variant, genome build issues, SNP identifier error, trait mapping vs genomic mapping

How are SNP-SNP interactions and multi-SNP haplotypes shown?​

The GWAS Catalog represents interactions and haplotypes using clear separators so that you can match each SNP’s details across fields.

  • SNP-SNP interactions: Elements are separated by x. Example: rs1336472-A x rs4715555-G shows the rsID and allele for each SNP. The same separator is used for chromosome locations (1p31.3 x 6p12.1) and variant consequences (3_prime_UTR_variant x upstream_gene_variant).

  • Multi-SNP haplotypes: Elements are separated by ;. Example:

  • rs17310467-?; rs6088735-?; rs6060278-?; rs867186-?

  • MYH7B; EDEM2 - PROCR; EDEM2 - PROCR; PROCR Each position corresponds across columns — the first rsID matches the first mapped gene, first location, and so on.

Mapped genes in the table and downloads: The Catalog can map the component SNPs in a haplotype individually. In the Associations table, each SNP’s mapped-gene value follows the SNP order; semicolons separate the values.

For an intergenic SNP, the nearest upstream and downstream genes are joined by - (for example, EDEM2 - PROCR). The MAPPED GENE(S) full-download field uses the same hyphen convention. The table displays gene labels as clickable links; downloads contain plain text.

In REST API v2: Association responses return mapped gene names in the mapped_genes array, not the table’s semicolon-separated display string.

Author-reported genes are separate under author_reported_genes in the association’s loci response. The REST API guide explains how gene filters select mapped or extended gene sets.

Mapped gene and position information is included in download files, but complex data such as upstream/downstream gene IDs and distances are excluded to keep the spreadsheets usable. For large haplotypes, this avoids confusing mismatches where some SNPs are intragenic and others intergenic.

Related search terms: SNP interaction display, haplotype format, rsID separator, multi-SNP annotation, x separator, semicolon separator, mapped gene alignment, MAPPED GENE(S), intergenic gene separator, mapped genes API, author-reported genes

What does the "Most severe consequence" or "CONTEXT" field mean?​

This field shows the predicted most severe functional effect of a variant, based on Ensembl’s consequence ranking.

A variant can affect multiple transcripts in different ways (for example, intronic in one transcript and missense in another). The Catalog only reports the single most severe consequence across all transcripts.

You can find definitions of each consequence term and their severity ranking in Ensembl’s documentation: https://www.ensembl.org/info/genome/variation/prediction/predicted_data.html#consequences

Related search terms: most severe consequence, variant effect, VEP, Ensembl consequence ranking, transcript annotation, functional effect

Why does the risk allele sometimes not match the reference genome?​

Risk alleles in the GWAS Catalog are taken exactly as reported in the original publication.

Occasionally, the curated risk allele does not match either allele in the current reference genome assembly. This can happen for several reasons, most commonly strand flipping between genome builds (for example, if the study was reported on an older build).

Because authors do not always report which genome build they used, the Catalog cannot systematically correct for these discrepancies.

For a detailed discussion of this issue, see: Sheng et al. 2022

Related search terms: risk allele mismatch, strand flipping, genome build differences, allele orientation, reference genome discrepancy

F. Population descriptors​

How are population descriptors provided in the GWAS Catalog?​

The GWAS Catalog represents sample metadata including population descriptors using a standardised framework:

  • Detailed sample description — captures the author-provided description of each sample

  • Ancestry category label — assigns each sample to a controlled vocabulary (e.g. European, East Asian) to make data easier to group and compare across studies.

This two-level approach allows both detailed interpretation and high-level filtering of studies.

For more information, see the GWAS Catalog’s Population Descriptors documentation.

Related search terms: population descriptors, ancestry category, sample metadata, population standardisation, controlled vocabulary

How do I search for population descriptors in the GWAS Catalog?​

You can search for specific population descriptors using the Studies table on any Trait, Publication, Gene, or Variant page, via the Discovery/Replication Sample Number (default) or Discovery/Replication Sample Description columns (enable these via Column visibility).

For more information, see the GWAS Catalog’s Population Descriptors documentation.

Related search terms: population search, ancestry search, discovery sample, replication sample, column visibility, sample metadata

Can I find all associations with a particular ancestry label?​

The GWAS Catalog website does not currently provide a direct way to view all associations for a given ancestry label.

You can:

  • Use the REST API to filter and retrieve associations by ancestry.

  • Download the full ancestry metadata file (includes country of recruitment and additional sample information) from the GWAS Catalog download page.

  • See the file header descriptions for details about the fields.

Related search terms: ancestry label search, population filter, ancestry download file, REST API ancestry, sample metadata

What does the "Pre-2011 ancestry not double-curated" flag mean?​

This flag indicates that the ancestry metadata for that study was curated before 2011 and has not been reviewed by a second curator.

Because of this, the information may not fully follow the strict, standardised approach used for more recent studies. The data is still publicly available but should be interpreted with this limitation in mind.

Related search terms: population descriptors, curation quality, metadata review

Why is detailed sample metadata in a separate "Ancestry data" spreadsheet to the rest of the study-level information?​

Each GWAS study usually has multiple sets of sample metadata — for example, one for the discovery stage and another for the replication stage — and some studies have several entries per stage.

Because this information is multi-dimensional, it cannot be easily represented in a single row of the main studies file. To keep the data clear and usable, each ancestry label is listed on its own row in a separate spreadsheet called the "Ancestry data" file.

Related search terms: ancestry spreadsheet, sample metadata, discovery stage, replication stage, ancestry file format, multi-dimensional data

Does the GWAS Catalog curate cohorts? What is included in the COHORT field?​

Yes. For curated studies, the COHORT field lists the discovery-stage (genome-wide) cohorts. For published studies, cohort information is available in the Studies download files and REST API. It is not displayed on the interactive GWAS Catalog pages. Cohort names are extracted from the literature and matched to a predefined list shared with the PGS Catalog.

Key details:

  • The predefined list was first compiled from Mills & Rahal (2019): https://doi.org/10.1038/s42003-018-0261-x.

  • A downloadable list of cohort abbreviations and full names is available here: https://ftp.ebi.ac.uk/pub/databases/spot/pgs/metadata/pgs_all_metadata_cohorts.csv.

  • Because the list is shared with the PGS Catalog, some cohorts may not currently appear in GWAS studies.

  • If a cohort is not in the predefined list or is not clearly described, the COHORT field will show “other.”

  • If no cohort was reported in the paper, the COHORT field will show “NR.”

  • Older studies (curated before ~2020) may have an empty COHORT field, since no attempt was made to extract the cohort at that time.

Note: Cohorts listed in unpublished download files have not yet undergone curation and may not exactly match the predefined list.

Related search terms: curate cohorts, cohort field, GWAS cohorts, discovery stage, PGS Catalog, cohort abbreviations, Studies download, where to find cohort information, NR, other, missing cohort info

G. Programmatic access​

How do I use the GWAS Catalog REST API?​

The GWAS Catalog REST API allows programmatic access to all Catalog data, including publications, studies, associations, traits, and variants.

Get started:

The API supports queries by study accession (GCST), PubMed ID, trait, variant (rsID), and more. Results are returned in JSON format, making them easy to integrate into pipelines or analyses.

Related search terms: GWAS API, REST API, programmatic access, JSON output

Is there a Summary Statistics API?​

No, we do not currently support API query access to GWAS Catalog summary statistics. The previously available API was retired in Aug 2026.

We are working towards developing a new database to support the tremendous growth in summary statistics data.

Related search terms: Summary Statistics API, deprecated API, programmatic access, summary statistics database

H. Summary statistics​

What are summary statistics?​

Summary statistics are the complete set of association results from a genome-wide association study (GWAS). They include every variant tested in the study, along with effect sizes, p-values, standard errors, and other metadata — not just the top significant hits.

The GWAS Catalog curates the most significant associations (p < 1×10⁻⁵) for display in the website tables, but we also make available the full summary statistics files where possible. These provide a comprehensive view of the study and are essential for downstream analyses such as meta-analysis, fine-mapping, and polygenic score development.

Related search terms: summary statistics, full GWAS results download

How do I find out which publications have full summary statistics available?​

Studies with available summary statistics are marked with an icon in the results of the homepage search, as well as indicated on publication and study pages.

You can also browse the complete list of studies (published and unpublished) with available summary statistics on the Summary Statistics page.

Files are usually released as soon as possible — sometimes even before the study is included in the Catalog’s regular biweekly data release. If a publication states that summary statistics are available but you cannot find them in the list or on the FTP site, contact gwas-info@ebi.ac.uk for more assistance.

Related search terms: summary statistics availability, studies with summary stats, FTP site, unpublished data

How do I access summary statistics?​

All summary statistics files are stored on our FTP site, which is updated nightly with any newly ingested data. Direct links to FTP files also appear on individual study pages when available.

For more details, visit our Summary Statistics pages.

Related search terms: summary statistics access, FTP site, nightly update, harmonised data

How should I cite summary statistics downloaded from the GWAS Catalog?​

When using summary statistics, please include:

  • The GWAS Catalog study accession (e.g., GCST007240)

  • A reference to the GWAS Catalog

  • The date you downloaded the file

  • If applicable, a citation to the original publication

Example citation: “Summary statistics were downloaded from the NHGRI-EBI GWAS Catalog (Cerezo et al., 2025) on 01/11/2025 for study GCST007240 (Riveros-McKay et al., 2019).”

Related search terms: summary statistics citation, GCST accession, download date, publication reference, citing GWAS data

What are harmonised summary statistics?​

Harmonised summary statistics are summary statistics files that have been processed to follow a consistent format and set of field names, regardless of how they were originally submitted.

Harmonisation ensures that:

  • Column names and file structure are standardised.

  • Key fields such as rsID, chromosome, position, effect allele, other allele, effect size, and p-value are present and consistently formatted.

  • Data can be easily combined across multiple studies for meta-analysis or downstream applications.

These harmonised files are available through the GWAS Catalog FTP site when processing is complete.

More details can be found on the Summary statistics documentation page.

Related search terms: standardised summary stats, harmonised data, file format, consistent columns, data cleaning

What is the summary statistics metadata YAML file?​

Each summary statistics dataset has a summary statistics file and a summary statistics metadata YAML file.

The summary statistics metadata YAML file describes the summary statistics dataset and is separate from the full GWAS Catalog download files.

The YAML file records dataset-level metadata such as the study accession, genome assembly, trait, sample information and whether the summary statistics are harmonised.

Related search terms: summary statistics metadata YAML, YAML file, metadata file, summary statistics dataset

What is the difference between submitted and harmonised summary statistics?​

Submitted summary statistics are the original files provided by study authors.

Harmonised summary statistics are standardised by the GWAS Catalog so that fields are represented consistently across studies. Harmonisation supports consistent file structure, variant representation and forward-strand allele orientation where possible. Harmonisation does not change the reported association statistics.

Related search terms: submitted summary statistics, harmonised summary statistics, standardised files, forward strand, file format

Where can I find the genome assembly used for a summary statistics dataset?​

The genome assembly is always displayed for GWAS Catalog summary statistics.

For summary statistics submitted since 2020, genome build is mandatory and is recorded in the summary statistics metadata YAML file.

For older submissions, genome build may have been inferred during curation. It is still displayed either in the YAML file or, for legacy studies, in the summary statistics filename.

Related search terms: genome build, genome assembly, GRCh37, GRCh38, hg19, hg38, YAML file, filename

What are the privacy risks of sharing summary statistics?​

Current consensus in the genetics community is that the benefits of openly sharing summary statistics outweigh the risks to participant privacy.

A 2008 study by Homer et al. (PMID:18769715) showed it was theoretically possible to detect whether a specific individual participated in a study if:

  • Their personal genotype data was publicly available, and

  • Summary-level allele frequencies from the study were also available.

Since then, multiple studies and reviews (including Craig et al., 2011, PMID:21921928) have assessed the risks and concluded that the risk is low.

The NIH supports open sharing of summary statistics, including allele frequencies, as stated in its guidance:

Related search terms: privacy risk, participant re-identification, allele frequency privacy, NIH guidance, open data sharing

Why do some datasets have a CC0 license mark?​

Since March 2021, all new submissions to the GWAS Catalog require data contributors to agree to share their data under the CC0 license.

CC0 dedicates the data to the public domain, allowing unrestricted reuse and removing any ambiguity about downstream use.

Key points:

  • Data submitted before March 2021 remains under the EBI Terms of Use, which already allow open reuse but are less explicit.

  • A small number of datasets may have other license terms — check the license listed on the study page or the Summary Statistics page.

  • Always cite the original data source when using summary statistics in a publication.

For questions about licensing, email gwas-info@ebi.ac.uk.

Related search terms: CC0, data license, public domain, summary statistics license, EBI terms of use, data sharing policy

I. Submitting summary statistics​

How can I submit summary statistics to the GWAS Catalog?​

The GWAS Catalog accepts summary statistics for both published and unpublished GWAS.

Summary statistics may be:

  • Uploaded by our team if they are openly available as supplementary files or via a public link, and meet our mandatory requirements.

  • Submitted directly by authors through the GWAS Catalog submission page.

Detailed instructions and formatting requirements are available in our submission documentation. All submitted studies must meet the Catalog’s eligibility criteria before summary statistics can be released.

Related search terms: summary statistics submission, unpublished GWAS, deposition page, file format requirements, eligibility criteria

How should summary statistics be formatted for submission?​

Summary statistics must follow the GWAS Catalog’s standardised file format.

See the full specification and access the validation tool here: /docs/summary-statistics-format

The validator will check your file structure, required fields, and formatting before submission to ensure compatibility with the Catalog.

Related search terms: summary statistics format, file specification, validation tool, submission requirements, data standardisation

Can I submit Z-scores as an effect size?​

Yes, but only when no other effect size measure is available, such as beta, odds ratio or hazard ratio.

This applies to cases such as certain meta-analyses or methods such as MOSTest.

If you submit a Z-score, it must correspond to the same statistical test used for the reported p-value, and the standard_error field must be #NA.

Related search terms: Z-score, z-score, beta, odds ratio, hazard ratio, standard_error, MOSTest, summary statistics format

Can I submit copy number variants (CNVs), structural variants or gene-based GWAS?​

Yes. The GWAS Catalog accepts some non-SNP submissions, including CNVs, structural variants and gene-based GWAS.

Current guidance and submission instructions are available here: https://www.ebi.ac.uk/gwas/apps/beyond-snps/how-to-submit/

Related search terms: CNV submission, structural variants, gene-based GWAS, beyond SNPs, non-SNP submission

Can I submit polygenic scores?​

No. Polygenic scores should be submitted to the PGS Catalog, not the GWAS Catalog.

Related search terms: polygenic score, PGS Catalog, submission, score models

Got a question that isn’t answered here?​

Email us at gwas-info@ebi.ac.uk.

Last updated: 07 August 2026