Google DeepMind Unveils AlphaGenome Atlas: Precomputed Molecular Effect Predictions and AVI Scores Across 9

Google DeepMind has released the AlphaGenome Atlas, a comprehensive catalogue of precomputed predictions detailing the molecular effects of every possible single-nucleotide variant in the human genome—amounting to nearly 9 billion individual letter changes. This resource introduces the AlphaGenome Variant Impact (AVI) score, a single metric that ranks variants by their predicted impact, alongside per-variant feature attributions and a genome-wide motif collection. As of 2026, the Atlas is freely accessible for academic research through a dedicated web portal, the AlphaGenome API, and as an integrated skill within Google Antigravity, reflecting the growing trend of making advanced genomic AI tools available to the broader scientific community.


Deployment and Accessibility

The Atlas is currently queryable for non-commercial research via the portal and API, with commercial access on Google Cloud expected to follow soon. The underlying AlphaGenome model is already available for academic use on GitHub and for commercial deployment through Model Garden on Google Cloud, enabling both researchers and enterprises to leverage this advanced tool for genomic analysis and interpretation.

via MarkTechPost

Related