Google’s AlphaGenome 2 zooms in on the cellular roots of disease
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. AlphaGenome 2 is the successor […]
Google DeepMind has unveiled an advanced AI model called AlphaGenome 2, designed to uncover how changes in our DNA affect different cells. This successor to AlphaGenome promises a more detailed view of genetic variations and their impact on individual cell types and states. By analyzing over 100 million cell profiles across 68 tissues, AlphaGenome 2 identifies nearly twice as many variants that can change gene expression compared to its predecessor.
The model's ability to predict effects on specific cells could reveal disease-relevant effects that might otherwise go unnoticed. AlphaGenome 2 also has the potential to make predictions about new cell types and biological states without requiring retraining. This specificity is crucial for researchers aiming to connect genetic variants identified in large populations to cellular-level effects, potentially leading to breakthroughs in understanding disease causes and treatment strategies.
The development of such models relies on advancements in capturing individual cells and their states at high resolution, enabling the creation of vast datasets necessary for training. While AlphaGenome 2 offers a more granular view than broader models, its computational demands make it impractical for producing genome-wide atlases.
Nonetheless, Google is making the model available for non-commercial research via the AlphaGenome API, with a private preview for commercial research. This breakthrough in AI-driven genetics is part of a growing trend among major tech companies and AI labs to apply their technology to major scientific and medical challenges.
Written by urgent.news from The Verge's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
