Google DeepMind Launches AlphaGenome 2 Genome Model

The updated DNA-sequence model predicts how genome variants change gene activity within specific cell types.

Updated on Nov. 11, 2026 in Biotech

Bold flat-color editorial illustration of twisted double-helix segments fanned apart in deep red and cream, showing genetic structure split into separate units.
Google DeepMind released AlphaGenome 2 on November 11, 2026, a DNA model that predicts how genome variants alter gene activity in individual cell types. AI Illustration. Upload story photo >

Google DeepMind released AlphaGenome 2 on November 11, 2026, an update to its DNA-sequence model. The new version predicts how genome variants alter gene activity in individual cell types rather than whole tissues.

Why it matters

The original AlphaGenome worked mostly at the level of tissues and bulk samples, which washes out signals from individual cell types. Genome-wide association studies have identified thousands of disease-linked variants without pinpointing the responsible cell.

AlphaGenome 2 reads one million DNA letters at a time and was trained on more than 100 million individual cell profiles drawn from 68 tissues in humans and mice. On the TenK10K dataset it identified causal variants 30% of the time, up from 16% for the first model.

The players

Google DeepMind

The Google-owned artificial intelligence research laboratory that developed the AlphaGenome models and has led advances in protein and genomic prediction.

The details

Non-commercial researchers can access AlphaGenome 2 through the AlphaGenome API, while commercial users receive a private preview. The model adapts directly to new single-cell data without retraining.

Timeline

  1. 2025: Original AlphaGenome released.

  2. November 11, 2026: AlphaGenome 2 released.

  3. November 30, 2026: Community spotlight program for API credits closes.

The Tech Race

AlphaGenome 2 extends the 2025 original AlphaGenome, which worked mainly at tissue and bulk-sample levels that obscured individual cell-type signals. The update moves the model toward cell-resolution genomics as AI labs compete to turn large genomic datasets into usable biological predictions.

Researchers can request access to AlphaGenome 2 through the AlphaGenome API, and commercial users can apply for a private preview. The community spotlight program offering API credits closes November 30, 2026.

The takeaway

Models that predict cell-level effects of genetic variants could help narrow the thousands of disease-linked variants that current studies have flagged but not explained. Watch for independent validation beyond the TenK10K benchmark before assuming the gains transfer to clinical work.

Further reading

For broader context on lab-driven advances in genetic research, explore our Biotech coverage.