On September 8, 2026, Google DeepMind announced "AlphaGenome Atlas," a platform designed to predict the effects of every single nucleotide substitution in the human genome. This platform provides predictive data for 9 billion changes across the entire genome, offering a comprehensive map of how genetic mutations influence molecular biology.
Product LaunchesGoogle DeepMindAlphaGenome Atlas
Google DeepMind Announces AlphaGenome Atlas to Predict Impacts of All Base Substitutions in the Human Genome
This article is a translation. Read the Japanese original
AlphaGenome Atlas was constructed by pre-computing predictions on a massive scale using the company's AI model, "AlphaGenome." Furthermore, it provides the "AlphaGenome Variant Impact (AVI) score," which combines the strengths of AlphaGenome with "AlphaMissense," a model used to predict the effects of protein mutations. By using this score, researchers can rapidly rank the impact of mutations and interpret their effects at the molecular level. The AVI score functions not only in protein-coding regions but also in non-coding regions (approximately 98% of the genome) that regulate gene activity.
This resource is currently available for free to academic researchers via a web portal and API. It has been reported that in collaborative research, the platform is being utilized to identify mutations related to rare diseases of unknown cause and to discover mutations in non-coding regions associated with common traits.
Sources
- AlphaGenome Atlas: A predictive map of every possible DNA letter change in the human genome （Google DeepMind Blog, 2026-09-08）