DeepMind unveils AlphaGenome 2 for cell-level genetic variant effects
Original title: Google’s AlphaGenome 2 zooms in on the cellular roots of disease
Trained on more than 100 million cell profiles, it is available today for non-commercial research through the AlphaGenome API.
Google DeepMind unveiled AlphaGenome 2, a successor to last year’s AlphaGenome model that predicts how genetic variants affect individual cell types and states. The company says it was trained on more than 100 million cell profiles across 68 tissues and identifies nearly twice as many expression-changing variants as its prior model. It can also predict unseen cell types without retraining, but DeepMind says the finer predictions cost more to compute. Non-commercial access is open via the AlphaGenome API, with a limited private preview for commercial research.
Unconfirmed: Performance claims, including nearly twice as many variants found, come from Google DeepMind.
- Google says the model covers 68 tissues and more than 100 million cell profiles.
- It is not replacing AlphaGenome, which still supports cheaper genome-wide predictions.
- DeepMind plans more training data, including poorly represented cell states.
- Future versions could use data from multiple species, Avsec said.
- Vendors
- Products
- AlphaGenome API
- Organizations
- Google DeepMind
- AI models
- AlphaGenome 2AlphaGenome
From the original article
3 min read at theverge.com
A graphic representing Google’s AlphaGenome Atlas. | Image: Google Researchers at Google DeepMind have unveiled a new AI model designed to unravel how changes in our DNA affect different cells, giving scientists a more detailed view of a notoriously difficult problem that could hold important clues to the causes of disease - and point the way to new treatments.
By Robert Hart · theverge.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.