AlphaGenome 2: Deciphering how cells read DNA to better understand disease
Predicting how DNA sequences and their variations impact specific human cell types and disease.
First seen by Oriel on .
AI summary
Google DeepMind introduces AlphaGenome 2, a conditional DNA sequence model that predicts cell-type-specific variant effects and identifies nearly twice as many expression-changing variants as its predecessor.
- Trained on over 107 million human and mouse cell profiles, it predicts gene expression, histone modifications, chromatin accessibility, and DNA methylation.
- The model conditions a million-letter DNA sequence on one tissue, cell type, or state, and can derive unseen states from scientists’ single-cell RNA-seq data.
- On TenK10K, AlphaGenome 2 identified 30% of causal variants versus 16% for AlphaGenome, and 43% of directional gene effects versus 23%.
- It can infer unmeasured properties across modalities and generalize to disease or vaccination states from snapshots containing only a few dozen cells, without retraining.
- AlphaGenome 2 is available through the AlphaGenome API for non-commercial research and private-preview commercial research, but is not validated for clinical or diagnostic use.
#genomics #single-cell biology #human genetics #ai models
Generated by gpt-5.6-sol. Check the original for details.
Aliases
| Source | External ID | URL |
|---|---|---|
| Google DeepMind Blog | https://deepmind.google/blog/alphagenome-2-deciphering-how-cells-read-dna-to-better-understand-disease/ | https://deepmind.google/blog/alphagenome-2-deciphering-how-cells-read-dna-to-better-understand-disease/ Read discussion |