Google DeepMind 发布 AlphaGenome 2，可预测 DNA 变异对特定细胞类型的影响

Beyond the model itself, the episode shows how AI variant interpretation is moving from research demos toward clinical diagnostic workflows for rare diseases. Genomicists, computational biologists and rare disease patient foundations can use it to understand what the model does and where it falls short, straight from its creators. Compared with academic preprints, the conversational format makes the transformer-based architecture accessible to non-specialists.

Official accountGoogle DeepMind@GoogleDeepMindSource

The practical gain is roughly doubled sensitivity to non-coding regulatory variants, which is where most GWAS hits and rare-disease candidates live, making the model directly useful for variant prioritisation. Genomics researchers, clinical genetics teams, and biotech pipelines evaluating open genome models can fold it into existing annotation workflows. Teams working with tissue-specific datasets should check whether their cell types fall within the 68 covered tissues.

Official accountGoogle DeepMind@GoogleDeepMindSource

Benchmark performance nearly doubled against its predecessor, which matters more for drug discovery teams than raw model size comparisons. Genomics researchers and biotech teams can use variant-to-cell-type predictions to prioritise which gene targets to validate experimentally. Compared with frontier language model releases dominating headlines, this shows AI value concentrating in scientific tooling, a direction enterprise R&D leaders evaluating AI investments should watch closely.

kimmonismus@kimmonismusSource

The move from tissue-level averages to single-cell-type resolution changes what genetics labs can do with variant interpretation: rare-disease teams can now pinpoint which specific cell type a variant disrupts rather than relying on broad tissue signals. Computational biologists and genomics researchers can start testing it immediately via the free AlphaGenome API for non-commercial work. Compared with prior variant-effect predictors, the conditional architecture and cross-modal imputation reduce the need to run every assay before planning follow-up experiments.