DeepMind launches AlphaGenome 2 for deeper single-cell genome analysis
Google DeepMind announced AlphaGenome 2, a next-generation genome AI model designed to predict how DNA sequences and variants affect specific human cell types and states. [[fact:fact-ty5a06w44]] The model uses more than 107 million single-cell atlases from human and mouse samples, spanning 68 tissue types. [[fact:fact-dwead22wo]] Compared with its predecessor, AlphaGenome 2 identifies nearly twice as many genetic variants related to gene expression. [[fact:fact-lrv4154aq]] In a TenK10K dataset test covering 28 primary immune cell types, the model identified pathogenic causal variants at a rate of 30%, versus 16% for the predecessor fine-tuned on single-cell data. [[fact:fact-mx9g6i8ss]] AlphaGenome 2 is available for noncommercial research through the AlphaGenome API, while commercial research users can apply to a limited private preview program. [[fact:fact-k0nmphpbw]] DeepMind said the model is for scientific research only and has not been validated or approved for clinical or diagnostic use. [[fact:fact-py96tjk7p]]