We are aware of an issue in the gnomAD v4.0 exomes where well covered variants have lower than expected allele numbers. This issue will be fixed in the upcoming v4.1 release. For more information, see our write-up here.


The changelog contains a record of all changes made to gnomAD, small or large.

Fix gene page variant table export to csv

The CSV file containing the information in the variant table for a given gene has been corrected to display the proper counts for each genetic ancestry group.

Any user utilizing a CSV file generated from a v4 gene page should re-download the CSV file to ensure they have the correct data.

Fixed chromosome Y VCF for v4.0 exomes

The exome chromsome Y VCF that was originally released on November 1st, 2023 was a duplicate of the genome chromosome Y VCF. The duplicate has been replaced by the correct VCF as November 16th, 2023.

Filter ClinVar Variants by review status

The ClinVar variants track now includes an option to filter by review status. ClinVar assigns a review status to each variant, along with a corresponding number of gold stars, ranging from 0 to 4 stars.

Users will now be able to filter ClinVar variants based on review status by selecting one of the following options from a dropdown: 0-4 Stars, >=1 Stars, >=2 Stars, >=3 Stars, 4 Stars.

Displaying neighboring variants when searching the variant table

Previously, performing a search in the Variant table would remove all neighboring rows from view, leaving only the exact search results displayed. This often resulted in a loss of context regarding a variant’s position in the table and its surrounding variants.

Now, there is an option to enable the display of nearby variants. When “Display neighboring variants” is selected, users are able to search for specific variants and also view nearby variants in the table. As users search, the table automatically jumps to the first search result, which is highlighted.

Population frequency display modified

Previously, for a given population, when a variant had an allele number (AN) of 0, the Allele Frequency (AF) in the population frequencies table would be displayed as 0. This was misleading, as it implied understanding of the frequency of the variant, where there was none.

Now, in those cases, the AF is now displayed as a ’-’ character. This helps users distinguish at a glance when there is no frequency information known about a variant for a given population or subpopulation.

ClinVar track bugfixes

The ClinVar track on Gene pages had several bugs that were identified and fixed, resulting in a more stable and accurate experience for users.

Genome build added to dropdown menus

Previously, when selecting a gnomAD variant dataset in the search bar or in the dataset selection dropdown, the dataset’s reference genome version was not displayed.

With this change, the reference genome version (GRCh37 or GRCh38) is now visible next to the gnomAD variant dataset version, making it more clear to users which dataset to select.

Modify navigation bar links

The ordering of the links on the navigation bar has been updated to reflect the frequency of their usage. The order of the links is now: About, Team, Policies, Publications, Blog, Changelog, Downloads, Help/FAQ.

In addition, the link formerly titled “News” is now “Blog”, and the link formerly titled “Help” is now titled “Help/FAQ”.

Fix co-occurrence table rendering on v3 datasets

Previously, the variant co-occurrence table on the gene page incorrectly rendered on v3 datasets. This was unintentional and misleading, as the analysis was not performed on v3.

Now, the co-occurrence table only renders on v2 datasets, as intended.

Add All of Us link on variant page

A link to the All of Us (AoU) data browser has been added to the variant pages under the “External resources” section. This lets users quickly navigate to the same variant on the AoU data browser.

Change gene search results sorting

Previously, when a user searched for a gene, a gene that had an alternate symbol that closely matched the search term could appear first in the list of suggested gene symbols. This led to confusing behavior and the need to pay close attention to which results the search bar returned.

Now, any gene with a canonical symbol that matches the search term is prioritized in the list of search results. This results in an overall more intuitive user experience when searching for genes by symbol.

As an example, the search term ABCD previously would result in the gene CX3CL1 as the first search result, as ABCD-3 is an alternate symbol for CX3CL1. With this change, searching for ABCD now results in ABCD1 as the first search result, as it prioritizes canonical gene symbols.

Update to Elasticsearch 7

The gnomAD browser’s backend database has had a major version update applied. It is now running Elasticsearch 7 and was previously running Elasticsearch 6. This update has been tested thoroughly and should represent no change in functionality or service.

Update GraphQL API to TypeScript

The gnomAD browser’s public GraphQL API has had its language updated to TypeScript, from JavaScript. This update has been tested thoroughly and should represent no change in functionality or service.

Update publications page

The publications page has been updated to include a more up to date list of publications, as well as to update the citation format to denote which authors contributed equally.

With the update of the publications page, the citations topic on the help page has been removed. The citations topic URL now redirects to the updated publications page.

Add gene page link on variant page

Previously, users that wanted to go to the corresponding gene page, from a variant page, would scroll down to the “Variant Effect Predictor” section of the variant page, and click on one of the gene links. In cases where many distinct variant pages were open, users expressed interest in a more direct way to go to the gene page.

With this change, a button labeled “Gene” has been added to the header row of a variant page where there is a single canonical transcript. This button allows users to quickly navigate to the gene that the variant falls in. In the rare case where there is no single canonical transcript, there is no gene page button, and users should instead use the Variant Effect Predictor section.

In order to accomodate the extra button, the phrase Single Nucleotide Variant has been abbreviated to SNV on the corresponding variant pages.

Add favicon

A favicon has been added to the gnomAD browser. This will help users to distinguish which tabs are gnomAD at a glance. The favicon is styled like the coverage plot, showing coverage for both exomes and genomes.

Fix OS variant table flag display

Previously, if a variant had an Other Splice loss of function flag as a consequence, the variant table would incorrectly display both a Low Confidence flag (LC pLoF), as well as the Other Splice flag (OS pLoF).

Now, in those cases, table only displays the Other Splice flag.

New team page

A new team page centralizes information about the people who work on the gnomAD project. Pertinent information that was previously located on the About page has been moved to Teams, and additional details about the team have been added.

Zoom in on gene and transcript pages

The gene and transcript pages show variants located in or within 75 bases of a coding exon. For genes and transcripts with large coding regions, this can result in data in the tracks being difficult to see when it is condensed into a small area on the screen. Now, you can zoom in on genes and transcripts to focus on data in a specific region.

Navigate directly to structural variants

Previously, viewing structural variants in the browser required looking up a gene or region in gnomAD v2 and then switching to the structural variant dataset.

Now, structural variants can be directly navigated to by selecting “gnomAD SVs v2.1” in the search box on the home page.

Variant co-occurrence (phasing) information added

Today, we are pleased to announce the incorporation of variant co-occurrence (inferred phasing) information in the gnomAD v2 browser. Phase refers to the genetic relationship between a pair of variants; that is, whether the variants are on the same copy of the gene (cis) or on different copies of the gene (trans). We are releasing inferred phasing data for all pairs of variants within a gene where both variants have a global allele frequency in gnomAD exomes <5% and are either coding, flanking intronic (from position -1 to -3 in acceptor sites, and +1 to +8 in donor sites) or in the 5’/3’ UTRs. This encompasses 20,921,100 pairs of variants across 19,685 genes. We envision that this data will be of tremendous help to the medical genetics community in identifying and interpreting co-occurring variants in the context of recessive conditions.

See more information in the full blog post.

gnomAD VCFs removed from public bucket

Last year, as part of an effort to reduce costs, we shifted our public Hail-formatted tables into requester-pays buckets. In keeping with our commitment to make gnomAD data as free and accessible to the world as possible, we continued to make VCFs free to download, paying for these downloads ourselves.

Over the last year, we’ve worked with major cloud providers to make these VCFs available for free via their public dataset programs. We’ve decided, as a result, to remove gnomAD VCF files from our own bucket, and encourage users to make use of the (identical) VCF files hosted by our cloud partners.


On variant pages you can now quickly switch between genome builds and gnomAD versions using our new liftover feature. This is available on v2 and v3 variants.

gnomAD v3.1.1

gnomAD v3.1.1 contains some minor corrections and changes to the v3.1 data release. The major annotations, including allele count, allele number, and allele frequency, as well as variant filtering status, remain unchanged for the entire callset and for all subsets of the callset (except the change noted below about switching to Null instead of 0).

New help page

A new help page brings together content that was previously divided between different areas of the browser. Now all content, including frequently asked questions and help popups, can be searched in one place.

Change to population labels

In the population frequencies table for gnomAD v2.1.1 and gnomAD structural variants, we have changed the labels we use to classify individuals by chromosomal sex from “male” and “female” to “XY” and “XX,” respectively. More detail on the reason for this change can be found in the gnomAD v3.1 blog post.