Google DeepMind has introduced AlphaGenome Atlas, a platform containing predictions for the effects of 9 billion single-nucleotide variants — every single-letter change possible — in the human genome. Described as the most comprehensive catalogue of how genetic mutations affect molecular biology, it is available for academic research through an intuitive, free-to-use website portal.

DNA is the language of life, and mastering it is a grand challenge that could transform the ability to understand biology and treat disease. Progress has been limited by a fundamental problem: interpreting how genetic variations affect biology at the molecular level. With roughly 9 billion possible single-letter mutations in the human genome, testing each one in the lab is practically impossible. AI-driven prediction is therefore becoming a standard complement to laboratory work in genomics, much as it already has in protein structure research.

DeepMind had already made progress on this challenge with AlphaGenome, an artificial intelligence (AI) model that predicts how genetic variants impact biological processes. While AlphaGenome is useful for analyzing specific variants and has seen widespread research use, the team wanted to give researchers a big-picture view of variants across the entire genome. By precomputing AlphaGenome's predictions at scale, they created an easily accessible resource that vastly expands the model's reach. Just as an atlas is a collection of maps linking features of the land such as altitude and location, AlphaGenome Atlas charts the molecular effects of DNA variants across the genome.

To help scientists quickly find the most impactful genetic changes, DeepMind is also releasing the AlphaGenome Variant Impact (AVI) score. The AVI combines the strengths of AlphaGenome and AlphaMissense — the company's model for predicting the impact of protein-altering DNA variants, whose earlier catalogue of missense variant predictions was published in Science in 2023 — condensing both models' predictions into a single number, allowing researchers to rapidly rank variants and interpret their molecular effects at the same time.

Trusted external collaborators have already used AlphaGenome Atlas to identify and experimentally verify key variants in unsolved rare disease research and to find rare variants associated with common traits.

AlphaGenome Atlas is available through the website portal, the AlphaGenome API, and as a skill in Google Antigravity.

What AlphaGenome Atlas contains

AlphaGenome Atlas is a massive 1-petabyte dataset, more than 30 times larger than the AlphaFold Database. When the AlphaFold Database was expanded in 2022, the available 3D structure information grew from around 190K experimental structures to more than 200M structure predictions — covering nearly all catalogued proteins known to science. That database offered a portal usable by researchers with no coding experience, with intuitive visualizations that made large-scale protein structure analysis easier. It quickly became a crucial resource driving discoveries across the life sciences and continues to accelerate research in countless fields; the underlying AlphaFold work was later recognized with the 2024 Nobel Prize in Chemistry, awarded in part to DeepMind CEO Demis Hassabis and AlphaFold lead John Jumper.

In building AlphaGenome Atlas, DeepMind similarly aims to make predictions accessible and to give scientists an intuitive way to explore a vast dataset. The Atlas provides several interconnected resources that allow researchers to link variants directly to the functional DNA sequences they disrupt:

• Molecular effect predictions: thousands of molecular effect predictions for each variant, across multiple aspects of gene regulation, spanning hundreds of human and mouse cell types and tissues. This is the starting point for the other resources.
• AVI score: a single number describing the impact of each genetic variant.
• AVI feature attributions: each AVI score is linked to the distinct biological features driving it, such as the aspects of gene regulation predicted by AlphaGenome or the protein impact score from AlphaMissense.
• DNA sequence motifs: a comprehensive collection of over 2,500 recurrent DNA sequences — the "words" of the genome — and their locations.

Together, these resources support a wide range of genetic research tasks, from rapid variant ranking to deep dives into variant functions. Extensive community collaboration guided the platform's design.

The AVI score helps researchers rapidly score and rank variants based on their potential impact. Crucially, it works for both coding regions — the 2% of the genome that codes for proteins — and non-coding regions, the remaining 98% that orchestrates gene activity and houses most trait-associated variants. This matters because genome-wide association studies have historically located most disease-linked variants in non-coding regions, where their function has been hardest to interpret. DeepMind's testing shows the AVI score delivers best-in-class performance across many variant pathogenicity and rare disease benchmarks. To aid interpretation, AVI feature attributions highlight which molecular processes — such as RNA splicing or gene expression — are predicted to be most disrupted by each variant.

An overview of the workflow: (1) precomputed effects are generated genome-wide for over 9 billion single-nucleotide variants; (2) an allelic-resolution AVI score is derived for each variant, then decomposed into additive feature contributions across interpretable categories such as chromatin accessibility, splicing, and conservation; (3) the precomputed variant effects, AVI score, and AVI feature attributions are linked with a compendium of genome-wide de novo motifs, enabling high-resolution mechanistic insights into variant function.

Real-world impact: from rare diseases to population genetics and molecular biology

AlphaGenome Atlas provides a high-resolution, global view of the genome, and these large-scale predictions become most useful when applied to targeted research questions. Academic partners are already translating the data into biological insights linking genetic variation and disease.

Understanding unsolved rare diseases. A major hurdle in rare disease research is pinpointing the few causal variants hidden among thousands of candidates. In collaboration with the GREGoR Consortium, researchers applied the AVI score to prioritize these needle-in-a-haystack variants. When Laura Covill and Anne O'Donnell-Luria from the Broad Institute and their colleagues used the AVI score to prioritize variants driving a rare disease that had been overlooked in previous research, the team discovered a variant affecting the gene DNM1, which is strongly linked to epileptic encephalopathy.

Crucially, the AlphaGenome predictions underlying the AVI score showed exactly how the variant functioned: it created an incorrect splice site — a mistake in the cell's genetic instructions — leading to an abnormal extension of the resulting protein. Experimental screens validated the prediction and found nearby variants with similar effects, showing that Atlas is a powerful tool for understanding impactful genomic variation.

Mapping rare variants associated with protein levels and complex traits. Beyond individual rare diseases, AlphaGenome Atlas can help uncover the genetic architecture of common traits in the general population. Identifying which rare, non-coding variants are associated with a specific trait or disease is notoriously difficult because the sheer volume of harmless genetic changes creates statistical "background noise."

To test how the Atlas could improve detection of non-coding variants affecting human traits, Gareth Hawkes, a Medical Research Council fellow at the University of Exeter, applied AlphaGenome Atlas to whole-genome data from more than 54,000 UK Biobank participants, making these elusive signals more obvious. The UK Biobank, one of the world's largest biomedical databases, holds genetic and health data on roughly half a million participants and has become a cornerstone resource for population-scale genetics. By grouping rare variants based on their predicted molecular effects, Hawkes uncovered 22% more non-coding genetic associations that would otherwise have been undetectable in the statistical noise. This allowed him to pinpoint specific regulatory variants driving the abundance of critical proteins circulating in the human body, including PLA2G7 (linked to aging) and EGLN1 (a vital cellular oxygen sensor).

Taking the approach further, Hawkes used AlphaGenome Atlas to examine how hundreds of millions of non-coding variants in the UK Biobank might be linked to body mass index. By focusing on the 1% of non-coding variants that Atlas predicts to be most impactful, he identified 19 genetic regions that could help direct the next stage of targeted research into this trait.

Identifying the regulatory "words" of the genome. Atlas can also identify which recurring short sequences, or motifs, drive different molecular processes in different cell types for different genes. These motifs can provide key clues, such as locating binding sites of transcription factors (proteins that turn genes on and off) and offering additional interpretation of non-coding variants. Julia Zeitlinger and Melanie Weilert at the Stowers Institute for Medical Research used this resource, for example, to categorize which transcription factors only affect DNA accessibility versus which ones are also able to turn genes on and off.

Accelerating genomic discovery

With AlphaGenome Atlas, DeepMind is creating new layers of information intended to advance understanding of the human genetic code. The company views the resource as a baseline rather than an endpoint: as AI models like AlphaGenome improve, maps of the entire human genome will become increasingly comprehensive and precise.

AlphaGenome Atlas is powerful in isolation but also represents a step toward DeepMind's vision of a broad, unified solution for biologists. Its resources can be integrated into broader agentic systems, such as Google Antigravity, to enhance end-to-end scientific workflows.

DeepMind has made AlphaGenome Atlas accessible for non-commercial use through its website from launch day, with commercial availability on Google Cloud to follow soon. The AlphaGenome base model is already available for academic use on GitHub and via the AlphaGenome API, and for commercial use on Cloud via Model Garden.

Together, these tools are intended to enable researchers and industry partners to accelerate biological discovery: finding novel therapeutic targets, better understanding genetic disorders, and driving the next wave of targeted experimental validation.

Acknowledgements

DeepMind thanked its research collaborators at the University of Exeter, Broad Institute, Boston Children's Hospital, Stowers Institute for Medical Research, Harvard University, Memorial Sloan Kettering Cancer Center, Center for Genomic Medicine at Massachusetts General Hospital, and the University of Kansas Medical Center.

DeepMind noted that the information provided by AlphaGenome Atlas is not intended as a substitute for professional medical advice, diagnosis, or treatment, and does not constitute medical or other professional advice. AlphaGenome has not been validated for, and is not approved for, any clinical use.

Source: Google DeepMind Blog