← All stories
● Covered by 1 source · 1 reportMedium impact1 neutral

Google DeepMind's AlphaGenome 2 Predicts Genetic Variant Effects on Individual Cell Types

🔄 Updated 10d ago
New to BrevFeed? We gather this story from every outlet covering it into one summary — ranked by real-world impact, not just the latest headline — so you never miss what matters. What is BrevFeed? →

Key points

  • AlphaGenome 2 predicts genetic variant effects on individual cell types.
  • Model trained on over 100 million cell profiles across 68 tissues.
  • Identifies nearly double the gene expression-changing variants.
  • Can make predictions on new cell types without retraining.

AlphaGenome 2 Unveiled

Google DeepMind researchers have introduced AlphaGenome 2, an AI model designed to analyze how DNA changes impact different cells. This new model provides a more granular understanding of genetic variations, which is crucial for identifying disease causes and potential treatments. It builds upon the original AlphaGenome model released last year.

Predicting Cell-Specific Genetic Effects

While all human cells share the same DNA, gene expression varies significantly across different cell types and states. These variations mean a single genetic variant can have diverse effects depending on the cell. AlphaGenome 2 is specifically designed to capture these differences, predicting the impact of genetic variants on individual cell types and their current states. This capability offers a more detailed perspective than broader models, potentially uncovering subtle disease-related effects in specific cells.

Training and Performance Improvements

AlphaGenome 2 was trained using over 100 million individual cell profiles sourced from 68 different tissues. This extensive training data allows the model to transform static cell atlases into a predictive tool for human genetics. DeepMind reports that AlphaGenome 2 can identify almost twice as many variants that alter gene expression in cells compared to its previous state-of-the-art model.

Advanced Predictive Capabilities

A notable feature of AlphaGenome 2 is its ability to make predictions about new cell types and biological states without requiring retraining. The model can also pinpoint gaps in existing experimental data, which could guide scientists in planning future laboratory investigations. Researchers involved in the project highlight its significance in connecting human genetics with cell biology, enabling a deeper dive into genetic variants identified in large populations.

✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →

The daily brief

One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.

One email a day. Unsubscribe in one click, any time.

Today's brief

Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.

~4 min · 3 stories · Nov 21

▶ Play today's brief Listen on Spotify

New every morning, and the back catalogue is archived by date.

Reporting from

Google DeepMind has released AlphaGenome 2, an AI model that predicts how genetic mutations affect specific cell types and states. This model offers a more detailed view than previous versions, potentially revealing disease-relevant effects that were previously undetectable. AlphaGenome 2 identifies nearly twice as many gene expression-altering variants compared to its predecessor.