Google DeepMind unveils AlphaGenome Atlas mapping all 9 billion human DNA mutations

September 8, 2026 at 07:00 PM3 min

This digest was compiled by AI from multiple sources — links to the originals are below.

Google DeepMind on Tuesday unveiled AlphaGenome Atlas, an AI-generated catalogue predicting the molecular effects of all 9 billion possible single-letter changes to the human genome. The database, roughly 1 petabyte in size, is freely available for non-commercial use through a web portal, Antigravity, and AlphaGenome. Researchers can now rank variants and interpret their molecular effects at the same time.

Key Facts

AlphaGenome Atlas contains predictions for all 9 billion possible single-letter DNA substitutions in the human genome.
The dataset is roughly 1 petabyte in size and is freely available for non-commercial use.
Google is releasing a Variant Impact Score (AVI) to help researchers rank variants and interpret their molecular effects.
The atlas builds on AlphaGenome, an AI model DeepMind unveiled last year, and extends predictions across non-coding regions of the genome.
Pushmeet Kohli, DeepMind’s vice president for research, said the release helps complete the unfinished business of the Human Genome Project.

The Atlas

AlphaGenome Atlas is a precomputed catalogue of what each substitution of a single DNA base is likely to do to the machinery that switches genes on and off. The human genome contains roughly three billion letter pairs, and there are roughly nine billion potential single-letter substitutions. Atlas contains predictions for how each of these nine billion variants could affect the body at a molecular level, such as changing how much of a particular protein is produced. Google says scientists can explore these predictions through a web portal, as a skill in its agentic development platform Antigravity, and through its AlphaGenome interface.

Model and Scale

The project builds on AlphaGenome, an AI model DeepMind unveiled last year to help scientists identify the genetic drivers of disease, as well as AlphaMissense, an earlier tool focused on predicting which small mutations might alter proteins. Atlas goes much further, extending predictions across the genome, including the vast majority of stretches that do not directly code for proteins, but can instead control how genes behave. Ziga Avsec, DeepMind’s genomics lead, said turning AlphaGenome’s capabilities into a genome-wide catalog took time because the space is so big. AlphaGenome was trained using public databases of human and mouse genomes, allowing it to learn patterns between DNA changes and biological processes. Applying those predictions to billions of possible variants produced a massive dataset that Google says is roughly 1 petabyte in size.

Research Impact

The atlas could help to diagnose rare, unexplained diseases and uncover the hidden biology of common illnesses and biological traits, say researchers. It might even reveal some of the hidden rules by which DNA sequences control gene activity. Martin Kircher, a bioinformatician at the Max Delbrück Centre for Molecular Medicine in Berlin, said the atlas won’t replace experiments or, in the case of diagnosing disease, accounting for details of individual cases. Pushmeet Kohli, DeepMind’s vice president for research and head of its AI for science team, said this was the first time any researcher in the world could reach a comprehensive map of human genetic variation by simply opening a browser. Kohli framed the release as helping to complete the unfinished business of the Human Genome Project, which in 2003 succeeded in mapping the entire human DNA sequence.