Google DeepMind releases AlphaGenome 2 to predict cell-type-specific DNA variant effects
Overview
Google DeepMind releases AlphaGenome 2, a model that predicts how DNA variants affect specific cell types.
The company says it was trained on more than 100 million cell profiles across 68 tissues and finds causal variants in 30% of cases in the TenK10K immune-cell dataset, versus 16% for the original AlphaGenome. A MarkTechPost report says the 30% figure comes from the source's own benchmark at 90% precision.
The model is available for non-commercial research through the AlphaGenome API and Colab tutorials. Commercial research access is limited to a private preview. DeepMind says it is intended for research use only and has not been validated or approved for clinical or diagnostic use.
Written by AI from the articles below · updated Nov 12, 6:46 AM ET
Check the sources:
Developments
5 developments
- Nov 12, 6:31 AM ET · 1 articleGoogle DeepMind explains AlphaGenome 2's role in rare disease researchGoogle DeepMind: Google DeepMind explains how AlphaGenome 2 helps find rare disease genetic causes
- Nov 11, 1:00 PM ET · 1 articleGoogle DeepMind explains AlphaGenome 2 and AI-based genome analysisGoogle DeepMind · YouTube: Google DeepMind explains how AlphaGenome 2 reads non-coding human DNA
- Nov 11, 10:03 AM ET · 1 articleGoogle DeepMind describes AlphaGenome 2's cell-type mutation detectionGoogle DeepMind: Google DeepMind's AlphaGenome 2 spots cell-type-specific gene regulation mutations
- Nov 11, 10:03 AM ET · 2 articlesGoogle DeepMind introduces AlphaGenome 2Google DeepMind: Google DeepMind introduces AlphaGenome 2 to study cell-specific disease drivers
- Nov 11, 10:00 AM ET · 6 articlesGoogle DeepMind releases AlphaGenome 2 for non-commercial researchGoogle DeepMind: Google DeepMind releases AlphaGenome 2 to predict cell-type-specific DNA variant effects
Article timeline
The articles in this story. Times are ET.
Google DeepMind@GoogleDeepMindOfficialGoogle DeepMind explains how AlphaGenome 2 helps find rare disease genetic causesAIGoogle DeepMind's podcast explains how AlphaGenome 2 lets scientists test millions of DNA changes to find clues to rare diseases. The episode covers the model's development from transformers, genome mapping, and rare disease cases, according to the post's timecodes.

- MarkTechPostNewsGoogle DeepMind releases AlphaGenome 2 for cell-type-specific DNA variant effects
AIGoogle DeepMind has released AlphaGenome 2, a DNA model that predicts variant effects in specific cell types through a 1 Mb input window and a target cell state. The model is available through a non-commercial API, with weights promised after final publication. According to the source's own benchmarks, it recovers 30% of causal variants at 90% precision on TenK10K single-cell eQTLs, versus 16% for the fine-tuned predecessor.
- Google DeepMind · YouTubeOfficialGoogle DeepMind explains how AlphaGenome 2 reads non-coding human DNA
AIGoogle DeepMind's YouTube episode, hosted by Hannah Fry, explains AlphaGenome, the AlphaGenome Atlas, and AlphaGenome 2. It discusses how AI analyzes non-coding DNA down to single-cell resolution to pinpoint disease-causing mutations. The source says about 98% of human DNA does not build proteins but controls how cells work.
Demis Hassabis@demishassabisOfficialPickGoogle DeepMind announces AlphaGenome 2, a DNA sequence model for cell-specific variant effectsAIDemis Hassabis announces AlphaGenome 2, a new DNA sequence model that predicts the impact of DNA variants in different specific cell types. The model is available for academic research through an API at alphagenomedocs.com.
Sundar Pichai@sundarpichaiOfficialPickGoogle DeepMind releases AlphaGenome 2 for non-commercial researchAIGoogle DeepMind has released AlphaGenome 2 for non-commercial research through the AlphaGenome API, according to Sundar Pichai's post on X. The post links to a DeepMind blog that describes the model as a tool for understanding how cells read DNA to better understand disease.
Sundar Pichai@sundarpichaiOfficialGoogle DeepMind introduces AlphaGenome 2 to predict how DNA variants affect cell typesAIGoogle DeepMind has introduced AlphaGenome 2, a model that predicts how specialized cells such as heart muscle and immune cells interpret the same DNA. Sundar Pichai says it builds on AlphaGenome 1 and makes it easier for researchers to link genetic findings to disease in the cells where it unfolds.

Google DeepMind@GoogleDeepMindOfficialGoogle DeepMind releases AlphaGenome 2 for non-commercial researchAIGoogle DeepMind releases AlphaGenome 2, which predicts how the same DNA is read differently from cell to cell. It is available now for non-commercial research through an API and interactive Colab tutorials.

Google DeepMind@GoogleDeepMindOfficialGoogle DeepMind's AlphaGenome 2 spots cell-type-specific gene regulation mutationsAIGoogle DeepMind says AlphaGenome 2, trained on over 100M profiles across 68 tissues, analyzes one cell type at a time. It identifies nearly twice as many mutations that turn genes on and off as its previous model.

Google DeepMind@GoogleDeepMindOfficialGoogle DeepMind introduces AlphaGenome 2 to study cell-specific disease driversAIGoogle DeepMind introduces AlphaGenome 2, an AI model that examines specific cells to help scientists understand what drives disease. The post says the same DNA sequence is shared across heart, brain, and immune cells, yet a single-letter change can affect disease in some cells but not others.

- Google DeepMindOfficialPickGoogle DeepMind releases AlphaGenome 2 to predict cell-type-specific DNA variant effects
AIGoogle DeepMind announces AlphaGenome 2, a conditional model that predicts how DNA variants affect specific human cell types. It identifies nearly twice as many causal variants as the original model in the TenK10K immune-cell dataset, 30% versus 16%. The model is available today for non-commercial research via the AlphaGenome API and in private preview for commercial research, and it is intended for research use only, not clinical or diagnostic purposes.
- The Verge · AINewsGoogle DeepMind releases AlphaGenome 2 for cell-level genetic prediction
AIGoogle DeepMind has released AlphaGenome 2, a 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 gene-expression-changing variants as its predecessor. It is available for non-commercial research through the AlphaGenome API, with a limited private preview for commercial research.
Heat Trend
- Current comparable heat
- 36
- Peak comparable heat
- 54Nov 11, 4:00 PM ET
- Change in past 24 hours
- –
The trend compares only continuously observed participants, so its scope may be smaller than the current heat total. Hover or tap for hourly heat; use arrow keys with a keyboard.