Google DeepMind has launched the AlphaGenome Atlas, a comprehensive database that uses AI to predict the molecular effects of approximately 9 billion possible single-nucleotide variants in the human genome. This resource, which includes an AlphaGenome Variant Impact (AVI) score, aims to accelerate genetic disease research by providing precomputed predictions, saving researchers significant time compared to individual variant testing or on-demand model computations. The Atlas is currently available for non-commercial academic use via a web portal and API, with commercial access through Google Cloud expected soon. A related paper detailing the creation and performance of the Atlas has been published on bioRxiv. AI
IMPACT Accelerates genetic disease research by providing precomputed predictions for billions of DNA variants.
RANK_REASON Release of a large-scale biological dataset with AI-driven predictions, accompanied by a technical report.
Read on Mastodon — mastodon.social →
- AlphaGenome
- AlphaGenome Atlas
- bioRxiv
- Google.Cloud
- Google DeepMind
- Human Genome Project
- Isomorphic Labs
- Pushmeet Kohli
- U.K. Biobank
- AlphaFold Database
- AlphaGenome Variant Impact (AVI) score
- AlphaMissense
- Google Antigravity
- Model Garden
AI-generated summary · Google Gemini · from 10 sources. How we write summaries →