Google’s Atlas of the human genome could pave the way for new treatments

Google DeepMind has launched AlphaGenome Atlas, an AI platform containing a predictive map for every possible single-letter DNA change in the human genome. The human genome consists of about 3 billion base pairs (A, C, G, T), with roughly 9 billion potential single-letter substitutions. This new catalog predicts the molecular impact of each variant, such as changes in protein production, making it a comprehensive resource for understanding genetic mutations.

To help researchers navigate this vast dataset, Google is also releasing a Variant Impact Score (AVI) tool, which uses the company's other models to rank and interpret the effects of DNA changes. The project builds on previous tools like AlphaGenome and AlphaMissense, but extends predictions across the entire genome, including non-coding regions that regulate gene behavior.

The dataset, approximately 1 petabyte in size, is based on precomputed analyses and is available to researchers for non-commercial use via a web portal. It represents a significant step in Google's ongoing efforts to apply AI to scientific research, coming as DeepMind co-founder Demis Hassabis shifts focus to drug discovery at Isomorphic Labs. The initiative follows the success of AlphaFold, for which Hassabis and John Jumper won the 2024 Nobel Prize in Chemistry.

(Source：The Verge)