Google DeepMind releases gene prediction model AlphaGenome 2

Google DeepMind released AlphaGenome 2, a gene-prediction model trained on more than 100 million single-cell profiles spanning 68 tissues, designed to forecast how genetic variants affect different cell types and states. Google said the model identifies nearly twice as many variants likely to alter gene expression as prior state-of-the-art models, and can generalize to new cell types and biological states without retraining. The model is available to non-commercial researchers through the AlphaGenome API, with a limited private preview for commercial research. The release matters because better variant-effect prediction could sharpen research into the cellular origins of disease and speed target discovery. What to watch next is uptake through the API and whether the commercial preview expands.