The API has been redesigned based on user feedback. It provides simpler querying, additional endpoints, richer filtering, ontology-aware searches and access to more Catalog data.

If you currently use the v1 API in scripts, pipelines or applications, this post explains the main differences and what you should review when migrating.

Starting something new? Use v2. The v1 examples below are included only to help existing users find code that needs replacing.

Why a new API?​

The GWAS Catalog has grown considerably since the original REST API was developed. V1 provides access to studies, associations, variants and traits, but its design increasingly made complex queries difficult.

For example, v1 exposes separate search endpoints such as findByPubmedId, findByRsId and findByEfoTrait. Searching SNPs includes an endpoint called findIdsByLocationsChromosomeNameAndLocationsChromosomePositionBetween.

V2 takes a different approach. Resources can be queried directly using filters, and additional entities such as publications, genes and ancestries have their own endpoints. The redesign is intended to provide a more robust and scalable API while making complex queries easier to construct.

The API remains focused on the Catalog’s literature-curated top associations and associated metadata. A separate redevelopment is underway for API access to the full genome-wide summary-statistics collection.

At a glance: legacy v1 compared with recommended v2​

The v1 column is included only to help existing users find and replace legacy code. For new scripts, pipelines and applications, use v2.

One of the most significant changes is how queries are constructed.

Conceptually, a v1 query might look like this:

/gwas/rest/api/associations/search/findByRsId?rsId=rs123

In v2, search criteria are supplied as filters on the resource:

/gwas/rest/api/v2/associations?rs_id=rs123

The same principle applies across the API. Instead of learning which findBy... endpoint corresponds to a particular query, users can start with the resource they want and filter it. This also makes the API better suited to more complex queries involving multiple criteria.

More ways to filter studies​

V1 already provides several ways to find studies, including by PubMed ID, Catalog accession, disease trait, EFO trait and availability of full p-value data. V2 expands the available filtering substantially.

Study searches can include information such as:

• pubmed_id
• disease_trait
• efo_trait
• efo_id
• accession_id
• cohort
• ancestral_group
• mapped_gene

V2 also provides additional study properties, pagination and sorting. Questions that previously required several requests, additional processing or another Catalog data source can increasingly be expressed directly as API queries.

The v2 release also exposes additional data including cohorts, background traits and fuller free-text sample descriptions.

More first-class resources​

V1 is centred primarily around four major resources: studies, associations, SNPs and EFO traits.

For example, publication information and ancestry information are contained within the study representation. A v1 study includes publication metadata alongside an array of ancestry records.

V2 exposes more concepts as resources in their own right. These include publications, genes and ancestries, alongside studies, associations, variants and EFO traits.

For example:

/gwas/rest/api/v2/publications
/gwas/rest/api/v2/genes
/gwas/rest/api/v2/studies/{accession_id}/ancestries

This makes it easier to start a query from the entity relevant to your analysis rather than retrieving another resource simply because the information happens to be embedded within it.

Better ontology-aware trait searches​

GWAS Catalog traits are mapped to ontology terms, allowing related phenotypes to be represented consistently. V2 can use that ontology hierarchy when performing searches.

When querying by an EFO trait, users can choose between retrieving records annotated directly with the requested trait or also retrieving records annotated with more specific child traits. For example, a query for asthma can either return only records mapped directly to asthma, or additionally include more specific concepts such as status asthmaticus.

This behaviour is controlled using the child-trait search option:

show_child_traits=false

This returns direct annotations only. To include descendant traits, use:

show_child_traits=true

For precise queries, using an ontology identifier such as MONDO_0004979 rather than a text trait name is recommended.

An important change to gene searches​

Users migrating gene-based queries should pay particular attention to the definition of the gene set.

V2 provides two definitions of the gene set associated with Catalog variants.

With extended_geneset=false, the API uses genes to which the variant maps, together with the nearest upstream and downstream genes according to Ensembl annotation. This corresponds to the annotation displayed by the current GWAS Catalog web interface.

For example:

https://www.ebi.ac.uk/gwas/rest/api/v2/associations?mapped_gene=HBB&page=0&size=20

With extended_geneset=true, the API uses the broader set of Ensembl and RefSeq genes mapping upstream and downstream of each variant:

https://www.ebi.ac.uk/gwas/rest/api/v2/associations?mapped_gene=HBB&extended_geneset=true&page=0&size=20

The extended gene set corresponds to the annotation used by the v1 API. V2 defaults to the first option to avoid confusing discrepancies with search-page results unless the broader set is intentionally requested.

Therefore, simply recreating a v1 gene query in v2 may not produce the same results. If reproducing v1 gene-query behaviour is important to your analysis, use extended_geneset=true when supported by the endpoint.

Changes to variants​

The naming of the SNP resource has also been modernised.

V1 uses /singleNucleotidePolymorphisms and identifies individual SNPs using an rsID. The SNP representation includes genomic locations, functional class and genomic contexts, with links to associated studies and associations.

V2 retains variant and rsID-based querying, but uses the redesigned v2 resource structure and supports filters including rsID, genomic location, PubMed ID, chromosome and mapped gene.

Update both the endpoint URLs and any code that depends on the structure of v1 SNP responses.

Changes to response handling and projections​

V1 is explicitly a hypermedia API using HAL. Responses contain _links, and the v1 documentation advises users to follow these links instead of constructing URLs themselves. Collections additionally place records under _embedded.

V1 also makes extensive use of projections, including convenience representations such as associationByStudy, associationBySnp and associationByEfoTrait.

V2 instead provides explicitly documented resources, filters and response schemas through its OpenAPI reference. For applications built around v1, migration involves more than changing request URLs. Code that parses v1 HAL responses or depends on projections should also be reviewed.

Pagination and API usage​

Both versions paginate large result sets. V1 returns 20 records by default and documents page and size, with up to 500 records in a single request.

V2 also defaults to 20 records per page. Applications retrieving complete result sets should follow the next links returned by the API rather than assuming that the first response contains every result.

V2 additionally has a rate limit of 15 queries per second. When this is exceeded, subsequent calls are slowed. Pipelines making many requests should account for both pagination and throttling.

More Catalog data through the API​

The redesign is not purely technical. V2 exposes data that was previously unavailable through the API or required users to obtain it through the GWAS Catalog website and downloads.

This includes:

• Cohorts: discovery-stage cohort information.
• Background traits: traits used to describe the study context.
• Fuller sample descriptions: more complete descriptions of study samples.
• Terms and licensing: information to help users understand reuse.
• Integrated annotations: additional annotations from resources such as Ensembl.

New endpoints for publications, genes, genomic contexts and ancestries also allow previously complicated queries to be performed more directly. The result is an API that more closely represents the breadth of information now available within the Catalog.

What do I need to change?​

Before you switch over, check these nine things:

If you maintain a pipeline or application using v1, migration should be treated as a small API upgrade project, not just a URL substitution.

In particular, check the following:

1. Endpoint URLs: move requests from /gwas/rest/api/... to the appropriate /gwas/rest/api/v2/... resource.
2. Search calls: replace /search/findBy... endpoints with v2 query parameters.
3. Parameter names: many v2 parameters use names such as pubmed_id, accession_id, efo_trait and mapped_gene.
4. Response parsing: do not assume v1 HAL _embedded, _links or projection structures will be present in the same form.
5. Gene searches: consider extended_geneset=true if you need behaviour equivalent to v1.
6. Trait searches: decide whether descendant ontology traits should be included.
7. Pagination: ensure your application follows paginated results.
8. Request rates: keep the documented 15-query-per-second limit in mind.
9. Data model: consider whether new resources such as publications, genes or ancestries provide a simpler way to perform queries that previously started from studies or SNPs.

GWAS Catalog REST API v2 documentation​

Read the GWAS Catalog API documentation for the interactive reference, endpoints, schemas, parameters and syntax.

For papers without submissions, our small team of biocurators manually extracts reported associations and metadata from GWAS publications. This careful work takes time, and we currently have a backlog of papers awaiting curation.

We know how important it is for your data to reach the community quickly. To help accelerate inclusion of new data in the Catalog, we are asking study authors to share metadata and top associations in a standardised format.

Authors contacted by the GWAS team are encouraged to review the checklist, submission guidelines and template before preparing their data. Providing data in a consistent format helps us to process the submission more efficiently, improve data quality, and reduce manual review time. This helps studies become available to the community faster, increasing their visibility and accessibility through the Catalog. As we are currently encouraging only a specific type of submission from external contributors, being contacted doesn’t guarantee that the study will be eligible for inclusion at this stage. All submissions will be checked and QC’d by our curation team before release.

Researchers who have not been contacted by the team but would like to contribute are also welcome to get in touch via the helpdesk, and we will guide you through the next steps. Please note - this is a separate process from submitting full genome-wide summary statistics - please don’t try to use our submission portal. Of course, if you are happy to submit full genome-wide summary statistics, we’d be very happy for you to do that as well!

As always, if you have any questions, please reach out to our helpdesk gwas-info@ebi.ac.uk.

The GWAS Catalog recently passed an incredible milestone of >1m curated associations! This is a huge achievement for the genomics community. However, it presents challenges for some users in analysing the data, since the “All associations” file now has more than 1m rows.

You now have a choice of two formats for the v1 and v2 “all associations” files:

1. File now provided as a compressed .zip; data format unchanged.
2. Data split into multiple smaller files, and compressed together. This split file is divided by publication year, with a separate file for each year from 2023 onwards, and pre-2023 data in a single file.

The split file (2) facilitates analysis for users of Excel or other software which cannot handle >1m rows. Please note that data may continue to be added for previous years - you are advised to always work with the latest version of the complete file set.

The files can be downloaded from the ftp links listed on the file downloads page.

If you’re trying to download the file programmatically for processing, we advise you do so from our public FTP. The previously available endpoint ebi.ac.uk/gwas/api/search/downloads/full was intended for manual access and has been deprecated.

To access subsets of the data, e.g. all associations for a given trait (e.g. “cancer”), you’re advised to either:

• Use the website to navigate to your data type of interest and then click “download Catalog data” to download relevant associations
• Use our recently updated REST API for programmatic access

Want to receive updates like this directly to your inbox? Sign up for our low-volume mailing list by sending an e-mail to gwas-announce-join@ebi.ac.uk with subject heading "subscribe"

Supporting programmatic access to the GWAS Catalog’s curated dataset, the REST API has been completely redesigned based on user feedback, to facilitate a robust and resilient mechanism for delivering the growing volume of data. The new design aims to provide all the tools you need to perform complex queries and effectively integrate GWAS Catalog data services into your scientific pipelines, scripts, and web applications.

New endpoints (including publications, genes, genomic context, and ancestries), allow users to create previously complex queries in a dynamic and simple manner. More data is exposed such as cohorts, background traits, full free text sample description, terms/license and other fields, which were previously accessible only in the GUI and downloads. Ontology powered search functionality, more search and filtering features, and additional integrated data from other sources such as Ensembl help to provide a richer user experience.

We’ve also made technical improvements to increase resiliency and scaling, applying industry standard architectural standards.

Extensive documentation is available, including a detailed reference manual containing endpoints, schemas, parameters, syntax and more, and including a visual, interactive interface for exploring and testing the endpoints. If you’re new to programmatic analysis, you might like to check out our Jupyter notebook tutorial. To get started, please go to our documentation page where you’ll find links to all of these. The previous API version will be available alongside v2 until May 2026, so you’ll need to switch over your pipelines before then.

Thanks to all our testers and survey respondents, and a special thank you to our super-user Ramiro Magno, developer of the gwasrapidd package, for making this work possible!

As always, if you have any questions, please reach out to our helpdesk gwas-info@ebi.ac.uk.

Here’s a hint of what you can find inside:

The GWAS-SSF standard for full genome-wide summary statistics data and metadata has been fully implemented in our submission pipeline. More than 30k files have been submitted containing ALL mandatory data & metadata fields, on top of 50k files in the previous format. We’ve also scaled our harmonisation pipeline to keep up with the increased rate of data submission - now 92% of all summary statistics with the necessary data content have a harmonised version available.

We have seen a huge growth in molecular quantitative trait GWAS (proteomic, lipidomic & metabolomic measurements) and are working with our colleagues at the Experimental Factor Ontology to improve mapping to chemical entities.

New features enable faster page loading and allow you to easily find GxE studies in our website and download files, preparing the way for anticipated growth in GxE data is anticipated to grow as the scale of exposome data expands. Thanks to funding from the Human Ecosystems Transversal Theme at EMBL-EBI to enable this new feature!

The lack of diversity in genomics data continues to be an issue. We share our thoughts on how to enable monitoring of diversity and inclusion in GWAS studies while taking care that the language we use to describe populations does not lead to misuse. New documentation and terminology (labels instead of categories) guide our users towards correct interpretation. Our analysis found that only around 50% of GWAS samples reported in the literature are being defined by genetic similarity measures, of those less than half clearly described a reference panel.

UKB is still the most used cohort in GWAS, accounting for around 30% of all studies in the Catalog! Unfortunately 70% of these restrict analysis to samples with the European population label, losing a significant amount of diversity. New inclusion of cohort names in Catalog data now enables this analysis, and in future linking out to rich cohort metadata.

Finally, we are now recognised as a Global Biodata Core Resource and Elixir Core Data resource!

Read the full article, Cerezo et al 2024, open access at NAR.

Comments, feedback…? Contact us via gwas-info@ebi.ac.uk

We have identified key fields that should be included in the metadata for a comprehensive understanding, interpretation, and subsequent reuse of GWAS studies. Mandatory fields document essential study details and optional fields cater to specific study designs (such as case/control counts in case-control studies), and provide supplementary analysis information. You can find detailed information about these fields in Table 1 and in our recent preprint (Hayhurst et al 2023 https://www.biorxiv.org/content/10.1101/2022.07.15.500230v4).

In our latest release, we've introduced several updates to the GWAS metadata YAML file format. The YAML format, known it's user-friendliness, provides a combination of structured information and ease of use for humans and computers. The latest changes aim to enhance the human readability of the files with a clearer, more unambiguous, and user-friendly experience without compromising the file's compatibility with automated systems. Here’s an overview of the significant enhancements:

1. Reordering and Grouping of Fields:

• The fields within the YAML files have been reorganized and grouped by related topics, each marked with a group name starting with '#' (group names are listed in Table 1). This organization helps users to quickly locate specific fields by group name, such as '# Sample Information', making the file more navigable for humans. Importantly, this structural refinement does not affect the way machines parse and interpret the data.

2. Clear Indentation for Hierarchical Relationships:

• Indentation has been introduced to visually represent the relationships between different data fields. This visual aid improves the readability of the files, making it easier for users to discern the levels of data organization and relationships between various fields. In the meantime, this indentation does not impact the machine processing of the file.

3. Renaming 'data_last_modified' to 'date_metadata_last_modified':

• To provide clearer context, the 'data_last_modified' field has been renamed to 'date_metadata_last_modified'. This change emphasizes that the timestamp refers to the last modification date of the metadata, not the summary statistics data. It aids users in tracking any changes or updates in the metadata over time.

4. Addition of 'Ancestry Category' Field:

• We have introduced a new 'ancestry category' field in addition to the existing 'ancestry' field. This new feature provides a broad overview of the ancestry category that best describes the sample, allowing for a more generalized view of the sample's ancestry. This new field also ensures consistency and facilitates comparative analysis across different studies, in line with the rest of the Catalog.

These updates reflect our commitment to improving the user experience while ensuring the YAML files maintain their structural integrity for computer readability. Our focus continues to be on making GWAS metadata more accessible, interpretable, and useful for researchers.

Table 1. Metadata field definitions

Questions and feedback​

Questions or comments about this change? Please contact us as gwas-info@ebi.ac.uk.

There are two new studies downloads named as follows:

• gwas-catalog-studies_ontology-annotated_v1.0.2.1.tsv
• gwas-catalog-download-studies-v1.0.3.1.tsv

Versions 1.0.2 (gwas-catalog-studies_ontology-annotated.tsv) and 1.0.3 (gwas-catalog-download-studies-v1.0.3.tsv) of the studies downloads will be deprecated and cease to be generated on 1 December 2023.

The COHORT field describes the discovery stage (genome-wide) cohorts used in each study. Since we started piloting curation of cohort information in 2020, over 700 different discovery cohorts have been described in GWAS publications.

Cohort abbreviations from discovery stage GWAS either are extracted from the literature or supplied by sumitters, to match a predefined list which we share with the PGS Catalog. The initial list of common cohorts used in genetics studies that seeded these annotations is from Mills & Rahal, Communications Biology (2019), and has been expanded during our pilot phase. A full list of the abbreviations and corresponding full cohort names in use is available to download separately (https://ftp.ebi.ac.uk/pub/databases/spot/pgs/metadata/pgs_all_metadata_cohorts.csv). We share this list with the PGS Catalog, so you may find cohorts listed there that are not currently associated with a GWAS Catalog study. It's important to remember we only extract cohorts when they have been clearly and unambiguously named by authors, so a cohort may have been used in other publications which don't have annotation. Where a sample cohort in the literature was not already in the predefined list at the time of curation, or was ambiguously described, “other” will appear in our studies download file. Where a sample in the literature had no cohort reported, “NR” will appear in our studies download file. Consortium names are extracted if the component cohorts are not provided, or there are too many component cohorts to curate. Empty cohort fields appear for studies curated before the pilot work to extract this information began (~2020). We also accept submissions of unpublished data and these are listed in separate download files. Cohorts appearing in the unpublished download files are yet to undergo in-house curation and therefore may not exactly match against the predefined list.

Cohort data annotations for published studies are also now available via the REST API, accessible via the /studies and /studies/{accessionId} endpoints. The field is called cohort.

As well as cohort data, the new studies downloads also contain clear annotation of which studies have full genomewide summary statistics available for download in the column FULL SUMMARY STATISTICS with the ftp location available in the adjacent column SUMMARY STATS LOCATION.

We hope you make great use of these data! Questions? Feedback? Contact gwas-info@ebi.ac.uk

The biggest change we have made is to increase the loading speed of pages, to account for recent increases in data volume and ensure we are robust for future growth. Check out a high-level query like “cancer” for maximum effect!

We’ve also main some small changes to the layout of the data pages, with tabs for the different data facets (i.e. associations, studies, visualisations), so you don’t need to scroll to find your data of interest:

Interested to find studies with full genomewide summary statistics available for download? They now have their own tab, making it easier to see when summary statistics are available:

Interested to find the study with the biggest sample size for your trait of interest? Just use the sort option on the Discovery sample number columnn, which now sorts by the total number of samples in the Discovery stage, irrespective of ancestry groups:

Hope you enjoy using the new features! Please report any bugs to gwas-info@ebi.ac.uk or via GitHub

Lastly, watch this space for the next generation of the GWAS Diagram - we're working on it!

This is a guide to the main differences you will see.

The biggest change we have made is to increase the loading speed of pages, to account for recent increases in data volume and ensure we are robust for future growth. Check out a high-level query like “cancer” for maximum effect!

We’ve also changed the layout of the data pages, with tabs for the different data facets (i.e. associations, studies, visualisations), so you don’t need to scroll to find your data of interest:

Studies with full summary statistics available for download are now easier to find in their own tab:

Please try out the new site and let us know what you like, what you don’t like, if you find any bugs or if any of your favourite features are missing. You can send feedback via gwas-info@ebi.ac.uk or report bugs directly via GitHub

Lastly, note that links to the GWAS Diagram have been disabled in the labs site. We are working on the next generation of the Diagram and will share this with you soon!

From 12th April we are switching to our NEW summary statistics format, GWAS-SSF, an overview of which is described in our manuscript and github repository. There will be a short period of submission system downtime while we switch over to the new pipelines.

What does this mean for you?​

If you are a SUBMITTER of summary statistics, you will need to submit additional mandatory information, including effect size, allele frequency, and genomic location for each variant, and columns must be presented in a specific order. You are also encouraged to provide additional structured data (e.g. reference allele, imputation info score) and metadata (e.g. ancestry method, sex of participants) which will increase the usability of your data to others, and hopefully get you more citations.

If you are a CONSUMER of summary statistics, you will start to see changes to the way we make these available. All newly submitted summary statistics will include the additional fields described here, with metadata available in a text file (.yaml) stored in the same ftp location as the data file, as well as its existing availability via the GWAS Catalog website and downloads. Over the coming months, we will generate metadata files for all existing summary statistics.

Frequently asked questions​

For submitters:​

I’ve already started a submission and prepared my files in the old format. What should I do? You can submit files in the old format until April 11th. If you are unable to complete your submission before this deadline, please contact us via gwas-subs@ebi.ac.uk for assistance.

I submitted summary statistics in the old format a while back and they are under embargo or in pre-published state. Do I need to change anything? No. Your data will be made available in the old format (upon journal publication, if embargoed), and the metadata file will be prepared from the information you submitted.

I want to start preparing my submission now, but won’t be ready to submit till after April 12th. How can I find instructions on how to prepare the new format?

Detailed documentation on the new format is available as follows:

• Summary statistics data file preparation
• Submission instructions for summary statistics and sample metadata
• Submission instructions for summary statistics only (for publications with metadata already available in the GWAS Catalog).

An example new-style metadata template is available to download here A recorded webinar is available, describing the submission process and the new format.

Are there tools available to help me prepare data in GWAS-SSF format? You can find some useful tools to format your files and validate them prior to upload here PLINK 2.0 will shortly include an option to generate output in the accepted format, using the option --gwas-ssf.

I made a submission in the old format but would like to update it to include the new fields. Can I do this? At the moment our update process is manual and we don’t have capacity to update large numbers of files unless there is an error. We plan to improve this process in the future.

For consumers:​

How can I tell if summary statistics are in the old or new format? This is indicated in the metadata file. If a metadata file exists, and the field named “fileType” has the value “GWAS-SSF v1.0”, then the file is in the new GWAS-SSF format. Other values in this field indicate that it is not in the new GWAS-SSF format, and may be lacking mandatory fields. If no metadata file exists, the summary statistics are in the old format, or a non-standard format.

How can I combine or compare new and old format summary statistics? Data fields included in the old format (chromosome, base_pair_location, effect_allele, other_allele, effect_allele_frequency, odds_ratio, ci_lower, ci_upper, beta, standard_error) are the same in the new format. There are two exceptions: variant_id and p_value. In the old format, variant_id contained rsIDs, whereas in the new format rsID has a separate column rsid, and variant_ID contains a variant identifier in the format chr_bp_ref_alt. In the old format, p_value was a mandatory field and could only contain p-values between 0 and 1. In the new format, p-values may be either presented in the p_value field OR as negative log10 p-values, with the header neg_log_10_p_value, but always in column 7.

More questions? Get in touch: contact gwas-info@ebi.ac.uk for general queries or gwas-subs@ebi.ac.uk for submission support

We started on a process of consulting with the community that began with our workshop on summary statistics data and sharing in 2020. This was a wide ranging discussion over 2 days that you can read all about in our workshop proceedings. At the end of that process we had an outline of some mandatory fields for data and metadata, but refining these and ensuring they met all the major use cases would take a more focused discussion. Our working group on summary statistics data content and format, chaired by Ines Barroso, held 3 meetings over the course of 2021. We held a parallel working group to discuss diversity and privacy in data sharing (to be reported elsewhere) and incorporated their recommendations (for example on the sharing of allele frequency data) into our decision making process. Both the original workshop and the working group were designed to represent a wide range of stakeholders - from industry and academia, developers of other resources & tools, data generators and data consumers, research groups primarily performing wet-lab research and high-throughput bioinformatics users. Our discussions focused on the need to enforce key mandatory fields to maximise the reuse of the data, but also to recognise that setting the bar too high could discourage data sharing. Similarly, data need to be made accessible in a way that suits the high-throughput user who might be piping data in bulk into a computational pipeline, but also users without advanced bioinformatics skills who might be downloading a single file to investigate a locus of interest. We heard that preparing data for submission to databases can be a huge drain for under-resourced labs, so needs to be as straightforward and user-friendly as possible, and that interoperability between different databases (such as dbGaP, the GWAS Catalog and OpenGWAS) is important to reduce the workload for data generators. You can read the outcome of our discussions in our new manuscript. Behind the scenes, we are making the necessary changes to the GWAS Catalog infrastructure to validate new submissions according to the new standard and where possible, make all our existing summary statistics available in the new format. We are working on tools (together with PLINK and others) to make data preparation and formatting as easy as possible. In the meantime, we welcome feedback from the wider community on the standard and will collate this at the end of August prior to release of the new infrastructure.

The GWAS-SSF is suitable for array-based GWAS and single variant seqGWAS. As data sharing has become more widely accepted, there is appetite from the genomics community to share full p value datasets from gene-based analyses of sequencing data and CNVs. Whilst we don’t have a standard format for these yet, by gathering data we hope to go through a similar process to work towards a standard format. Please share your data, and let us know your thoughts!

For each study, we annotate the trait in two ways:

1. A reported trait that reflects the author’s description of the disease or phenotypic characteristic under investigation. This is a free text description and sometimes different studies might use a different wording to describe similar traits or to capture more nuanced distinctions. Reported traits can also include multiple component traits, depending on the study design.

2. One or more trait terms from the Experimental Factor Ontology (EFO) that represent phenotypic characteristics in a more standardised way. These terms make studies on similar traits easier for users to find and compare. For multifaceted traits, each component is represented by a separate term.

Annotating different study types​

A. Single-trait studies​

The vast majority (>80%) of studies in the GWAS Catalog only analyse a single trait. We annotate these studies with a single EFO term.

Some common examples include:

1. Standard case-control studies comparing individuals with a disease or phenotypic characteristic, versus controls individuals without that trait.

2. Quantitative studies looking at a single measurement

In these examples, any reported variants are clearly and straightforwardly associated with that single trait.

B. Multi-trait studies​

In about 10% of studies, there are multiple traits of interest that are analysed simultaneously. We annotate these studies with multiple EFO terms separated by a comma, indicating that any significant variants reported in the study are associated with both traits in some way.

Some common examples include:

1. Studies comparing individuals with two comorbid diseases (or other traits), versus control individuals who have neither disease. In the reported trait we write “Disease 1 and Disease 2”. In the trait we list multiple EFO terms separated by a comma.

2. Studies comparing individuals with either of two diseases (or other traits), versus control individuals who have neither disease. Often these are two traits that are hypothesised to have some common underlying genetic factors (pleiotropy). In the reported trait we write “Disease 1 or Disease 2”. In the trait we list multiple EFO terms separated by a comma.

In these examples, any significant variants reported are associated with either or both of the annotated traits. If a user is searching for associations with either trait term, then these results will come up in their search.

C. Studies with a background trait​

Finally, about 6% of studies analyse only one main trait of interest, but in the context of a background trait that is shared by all of the participants in the study.

Some examples include:

1. Studies comparing cases and controls of one disease, but only within a sample of people who have another disease. In the reported trait we write “Disease 1 in Disease 2”. In the past, we listed multiple EFO terms in the trait column, as for scenario B.

2. Quantitative studies analysing a measurement in the context of a disease. In the reported trait we write “Measurement in Disease”. In the past, we listed multiple EFO terms in the trait column, as for scenario B.

In these examples, any reported variants are associated with the main trait, but not with the background trait.

Room for improvement​

In the past, we annotated background trait studies (scenario C) in the same way as we have for multi-trait studies (scenario B): with multiple EFO terms listed together in the trait field, separated by commas.

This had some benefits:

• We were able to indicate that the study has something to do with the background trait - e.g. a particular association with allergic rhinitis may only hold true in asthmatics.

• Users could search for associations with a trait, as well as associations that are found in the context of that trait as a background characteristic. Both kinds of associations might be relevant to users working in a particular field.

However, there were some are also some disadvantages:

• It was not possible to tell that a particular EFO term was a background trait, without looking at the reported trait field. This complicated analysis particularly when accessing the data programmatically.

• When searching for a trait term (e.g. asthma), there was no easy way to distinguish which studies were direct associations with asthma, and which have asthma as the background trait, without reading the reported trait for each one.

• The overall study and association numbers provided for some traits could be misleading since they included studies and associations where the trait of interest was a background trait.

Changes to trait annotation​

To make our trait annotations more informative, we have added an additional background trait field to the GWAS Catalog database.

We have moved all EFO terms related to background traits to this new field, and removed them from the original trait field:

Changes to the website​

We have also updated our web interface in order to display the restructured background trait information.

Data tables​

New columns have been added to the Associations and Studies tables to clearly indicate the main and background traits.

Associations:

Studies:

The improved tables will be displayed on all Publication, Study, Trait, Variant, Gene and Region pages.

Study pages​

Each Study page also displays the main and background traits in separate fields in the Study Information panel.

Trait pages​

The Trait page has been updated so that background trait studies and associations are no longer shown by default. For example, the asthma page only shows associations with asthma itself, and not associations with others traits in asthma.

Shows associations with:

• asthma
• [other traits] in asthma

However, users can select the box to include background trait data if they wish:

Shows associations with:

• asthma
• [other traits] in asthma

Some trait terms may only be used in the GWAS Catalog to annotate background traits. These terms will continue to have their own Trait page, but no associations or studies will be displayed under the default view. For example, autoimmune pancreatitis type 1 (EFO_1000780) currently appears in the Catalog only as a background trait for one study (“Lachrymal/Salivary gland lesion in type 1 autoimmune pancreatitis”), so no associations with autoimmune pancreatitis type 1 are displayed by default:

The association plot on each Trait page will now display only main trait associations by default, but background trait data can be added by selecting the box:

Downloads​

In the full Catalog downloads, the MAPPED_TRAIT and MAPPED_TRAIT_URI columns will now only show EFO terms for main traits. Background trait terms have been removed from these columns.

In the newest version of the download (v1.0.3), new columns have been added for the background trait: MAPPED BACKGROUND TRAIT and MAPPED BACKGROUND TRAIT URI. This is not included in earlier versions of the download.

API​

For users of the GWAS Catalog API, searching for “associationByEfoTrait” will now return only main trait associations.

There is currently no background trait field available in the API, but we plan to add this feature in the future.

Expected scope of the changes​

We reviewed all of the studies in the Catalog and identified just over 1000 studies (about 6% of the total) with background traits that have now been moved to the new field. These studies contain around 10,000 associations (about 4% of all associations in the Catalog) which have also been reannotated.

The changes affect some trait terms more than others. Here are some of the terms that will see the greatest change in the number of annotated Associations:

Questions and feedback​

If you have any questions or comments about this change, please contact us as gwas-info@ebi.ac.uk.

For more than 10 years, the Catalog has aimed to make GWAS data FAIR (Findable, Accessible, Interoperable and Re-usable), while serving as a starting point for investigations to identify causal variants, calculate disease risk, understand disease mechanisms and establish targets for novel therapies.

Here we aim to illuminate progress that has been made to incorporate GWAS summary statistics into the GWAS Catalog, and reach out to the cancer genetics community in particular to promote submission of summary statistics for GWAS studies in cancer cohorts.

The GWAS Catalog summary statistics repository​

Three years ago, the GWAS Catalog and Open Targets started to work together to expand the Catalog’s scientific scope to include full p-value summary statistics (aggregate p-values and association data for every variant analysed in a genome-wide association study) in addition to the manually curated top associations.

Summary statistics provide more detailed GWAS results than a lead SNP table, which allows other scientists to better use the data for sophisticated downstream analysis, whilst maintaining patient privacy. For example, summary statistics are integrated into the Open Targets Genetics Portal, where they are used to narrow down candidate loci and prioritise new drug targets through PheWAS, colocalisation analysis, fine mapping and mendelian randomisation

[GWAS-catalog-statistics-1.png

The GWAS Catalog is now one of the largest, most visited and most frequently updated resources of freely available GWAS summary statistics. The repository includes summary statistics from 3,623 independent analyses (from 447 publications), accounting for a total of more than 22,000 datasets from a wide variety of traits. The Catalog users can easily access and download summary statistics from the GWAS Catalog FTP site or via a dedicated summary statistics API.

The increasing trend to share these datasets is reflected in the availability of summary statistics in the GWAS Catalog over time. There was also a significant increase in data downloads of summary statistics in 2020 compared to the previous year.

Why are few open access cancer GWAS summary statistics available?​

This expansion of the GWAS Catalog summary statistics repository, while promising, required a considerable outreach effort from the Catalog data team and the main stakeholders, including Open Targets. Interestingly, the rate of summary statistics data sharing noticeably differs among different genetics cohorts and research groups, with the lowest submission rate in cancer genetics.

For example 33% of epilepsy and 21% of diabetes summary statistics are publicly available, compared to only 7.5% for cancer. The trend slightly differs amongst different cancer groups, and we have observed a more positive trend for the few papers published in the last two years.

In a recent Twitter poll, we asked the community to tell us what they thought was the most important reason for the low rate of cancer summary statistics submission.

The results suggest that the main barriers to the sharing of summary statistics are:

1. Results are usually embargoed for use in future research,
2. Data privacy issues (e.g. patient confidentiality agreement), and
3. Lack of awareness/knowledge on the appropriate data repository.

We believe it is necessary for the genetics community to have a more comprehensive and transparent discussion about such barriers, starting from the privacy issues. As highlighted in the recently revised NIH genomic data sharing (GDS) policy, summary statistics do not include individual-level information and can empower researchers to determine which genomic variants potentially contribute to a disease

How can we promote sharing?​

As a part of a wider plan to identify and remove barriers to data sharing among different research communities, we have launched a dedicated public engagement campaign to motivate and encourage the genetics community to make their summary statistics publicly available – and make the difference in the field. In particular, following up from a pressing request from our users and stakeholders, we hope to understand the reasons why so few summary statistics are available for cancer studies. Open sharing of cancer GWAS data will lead the way to new, improved therapies and shed some light on the molecular mechanisms involved in such complex disease.

We call on researchers, journals, funders and charities to support our cause and spread our message. The Nature Journals Group are already proactively supporting an open access policy for their published cancer GWAS data. We hope to get more journals on board, including the top cancer journals.

In an aim to facilitate data sharing and interoperability, the GWAS Catalog team has recently released a web-based deposition interface to support scalable author submission of summary statistics and metadata from published and pre-published (submitted at the time of journal submission upon request from the reviewers) GWAS. The datasets are submitted in a standard format and harmonised against the latest genome build and forward strand. This facilitates downstream analysis and integration into resources like the Open Targets Genetics.

Please submit your summary statistics to the GWAS Catalog.

What do you think are the most important barriers to the open sharing of summary statistics in cancer studies? Please let us know your thoughts by completing this short survey.

You can also reach out to us via Twitter (@GWASCatalog) or email (gwas-info@ebi.ac.uk).

Together, we can make the difference.

The project is a collaboration between the National Human Genome Research Institute (NHGRI) in Bethesda, Maryland, USA, and the European Molecular Biology Laboratory – European Bioinformatics Institute (EMBL-EBI) in Hinxton, UK, where the majority of our curators and all of our software developers are based

At EMBL-EBI, we share the building with the Ensembl team and work very closely with them – which is very handy for both teams!

Our curators extract genotype-phenotype associations from all published GWAS studies, along with study information to allow you to interpret the data accurately. We make the data structured, searchable and visualisable, annotate it and integrate it with other resources including Ensembl. Each association with a standard variant identifier (rsID) enters our automated Ensembl mapping pipeline which provides the genomic annotation, i.e. the location on the chromosome and cytogenetic band as well as the mapped genes. We use this pipeline to check for errors as part of our quality control and display the information in our interface.

Our colleagues at Ensembl benefit from the GWAS Catalog as they import variant and phenotype data from it. More than 74,000 human variants associated with phenotype data from the GWAS catalog were included in the latest Ensembl release (e92); and the number is increasing continuously!

Over the last ten years, the GWAS Catalog has grown, from a single published GWAS on age-related macular degeneration, into the primary source of disease-related associations with genetic variants. Watch how the number of associations has increased over time in this video!

Today, the GWAS Catalog contains over 3,400 publications and over 62,000 unique SNP-trait associations. More than 10,000 visitors from around the globe access the GWAS Catalog per month. People using the resource have a wide range of goals – they include researchers wanting to narrow down or prioritise candidate loci, scientists investigating disease mechanisms, clinicians aiming to predict disease risk, pharma industry professionals improving the drug discovery process and anyone wanting to get the latest statistics on disease knowledge or summary data from particular populations of individuals.

The GWAS landscape has evolved over these ten years, with new developments in study design and genotyping technologies. More and more publications on GWAS come out each month. Often a publication contains more than one GWAS – sometimes there are over a thousand in a single paper! The number of individuals studied in a GWAS is also increasing, and so is the number of variants assayed, with bigger arrays and better imputation methods.

The nature of the traits studied is evolving too. Researchers are still publishing GWAS on major common diseases such as type 2 diabetes and breast cancer. However, these studies are now being published alongside studies on ever more specific traits. For example, searching the GWAS Catalog for “schizophrenia” brings up more than 50 different traits related to this disorder. These include sub-phenotypes such as specific symptoms in schizophrenia, endophenotypes such as brain imaging measurements, multi-trait analyses such as “schizophrenia and bipolar disorder”, drug response measurements as well as analyses of the interaction between genetic associations and environmental factors.

All of this means more data to curate. We are adapting our infrastructure constantly to improve curation efficiency so that we can keep up with the rate of data generation. At the same time, we need to maintain the accuracy of the resource and make sure it contains the most relevant and up-to-date research results.

Highlights over the last ten years include releasing a new GWAS Catalog website in 2015, mapping curated trait descriptions to ontology terms to enable enriched ontology-driven search capabilities, improving the interactive GWAS Catalog diagram and capturing ancestry more completely. Accurate characterisation of ancestry is essential to interpret human genomics data. Recently, we have developed a new framework for describing detailed ancestry information and recommendations for reporting ancestry systematically Morales et al. (2018), Genome Biology, 19:2.

So what will the future hold for the GWAS Catalog? We have just started hosting summary statistics for studies in the GWAS Catalog with support from the scientific community (the files contain the full set of p values for every SNP on the array used). In the future we plan to integrate these datasets with structured meta-data to make it even easier to access GWAS results. We also aim to expand our scope to include other types of association analyses, including targeted arrays and sequencing-based genotyping, which enable deeper interrogation of diseases of interest and inclusion of less common variants.

With these developments come many challenges, including how best to represent the increasing complexity of study designs, trait architecture and statistical analyses. But with the help of the authors of publications in the GWAS Catalog, the scientific community using our resource and our colleagues at Ensembl, we are excited about what the next ten years of GWAS will hold!

The NHGRI-EBI GWAS Catalog, an awesome public resource of human genotype-phenotype associations from published GWAS analyses, is one of the data sources for manually curated genetic evidence for the Open Targets Platform. The GWAS Catalog data is widely used to prioritise new candidate loci, predict disease risk and understand disease mechanisms, with a number of users and applications that is always increasing.

New and expanded GWAS Catalog to improve target identification​

I am happy to announce an expansion of the GWAS Catalog scientific scope to host non genome-wide studies. During its first 10 years of life, in fact, the GWAS Catalog has only included array studies with genome-wide coverage, therefore curation of exome and targeted array studies were outside its scientific scope. These data are however valuable to Open Targets’ scientific goals of target identification and prioritisation, so much so they decided to invest in a new project aiming to expand the scope of the Catalog to include targeted arrays, such as MetaboChip, ImmunoChip, and Exome array. The new GWAS will now allow variants associated with immunologic, metabolic and oncologic phenotypes to be included in the Open Targets Platform for the expansion of target identification and selection in these traits.

Selecting and curating non-genome-wide array studies​

As a first step, Open Targets selected almost 60 publications for inclusion in the GWAS Catalog. I then extracted a total of 120 new independent association studies and 823 SNP-trait associations from these publications. These are impressive numbers, and give you an idea of how many studies and SNP-trait associations (and potential targets) we were missing by not including targeted array studies in the GWAS Catalog!

The final step for me to do was to map all these traits, diseases and phenotypes to EFO terms, the ontology used by Open Targets to describe its diseases

Modifying the GWAS Catalog user interface​

The GWAS Catalog team have changed the way the user interface works to support searching, displaying, filtering and download of targeted and exome array studies. Try it out using "celiac disease", for example. The interface will now give you:

• Drop-down filter by genotyping technology to help you focus on your favourite array

• Small "target" icon (designed for us by the EMBL-EBI media team. Thanks for that!) in the study accession number column for all targeted array studies prioritised by Open Targets

• New "genotyping technology" shown for study, e.g. genome-wide genotyping array, exome-wide genotyping array or targeted genotyping array with an optional field for additional array information, for example ImmunoChip or MetaboChip

Stay tuned!​

This is an important milestone for Open Targets and the GWAS Catalog, and a lot more is yet to come. The next step for the collaboration between GWAS and Open Targets will be to develop a comprehensive database of all available GWAS full summary statistics stored in a common format and harmonised across studies to enable searching, easy comparison and downstream analysis. This effort will generate more comprehensive new raw genetic data accessible by the entire research community through the GWAS Catalog, the Open Targets Platform and additional resources the Open Targets team are working on.