AlphaGenome 2 predicts cell-specific DNA reading patterns
The interesting shift is that the model's value comes from context — the same variant can be scored across many cell types, enabling tissue-aware interpretation of disease-linked variants that single-context tools miss. Computational biologists, genetics labs and variant-interpretation teams can start experimenting immediately through the free Colab tutorials without infrastructure setup. Commercial users are excluded from the current licence, so biotech teams should treat it as a research preview and watch how DeepMind prices commercial access later.
Google DeepMind has released AlphaGenome 2, an updated version of its genomics model that predicts how the same DNA sequence is interpreted differently from cell to cell. The model is available now for non-commercial research through DeepMind's API, accompanied by interactive Colab tutorials for hands-on exploration.
The core capability is modelling cell-type-specific regulation: identical genetic sequences can produce different outcomes depending on which cell type is reading them, and AlphaGenome 2 predicts these context-dependent effects. This matters because most genetic variation linked to disease sits in regulatory regions of the genome rather than in protein-coding genes, and its impact depends heavily on the tissue or cell type involved.
For researchers, the release lowers the barrier to running large-scale variant-interpretation experiments without bespoke pipelines, though access is restricted to non-commercial use for now. The update positions DeepMind further into biology, building on the AlphaFold lineage and competing in a space where academic genomics groups and biotech startups are also building predictive genome models.
By predicting how the same DNA is read differently from cell to cell, AlphaGenome 2 opens up entirely new possibilities for understanding human genetics. It’s available now for non-commercial research via our API and interactive Colab tutorials. Find out more → goo.gle/4zikg3w📊 3 ⚡
This report is categorized under AITOP's "Models" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include AlphaGenome, Google DeepMind, genomics, multimodal.
Source attribution: curated from Google DeepMind (@GoogleDeepMind).
