Google’s AlphaGenome 2 zooms in on the cellular roots of disease

Google DeepMind has introduced AlphaGenome 2, an AI model that predicts how genetic variants affect different types and states of cells. By linking genetic changes to their cellular effects, it could help researchers identify disease mechanisms and guide the search for treatments.

Trained on more than 100 million cell profiles across 68 tissues, the model identifies nearly twice as many variants that alter gene expression as its predecessor, according to DeepMind. It can also make predictions for cell types and states it has not been retrained on and highlight gaps in experimental data. Its detailed analysis is more computationally expensive, so it will complement rather than replace AlphaGenome; access is available for non-commercial research, with a limited commercial preview.