LONDON, United Kingdom: Google DeepMind has released AlphaGenome Atlas, a searchable database containing predictions for more than 9 billion possible single-letter changes across the human genome.
The roughly 1-petabyte resource uses DeepMind’s AlphaGenome model to estimate how DNA variants may affect molecular processes in both protein-coding and non-coding regions.
Google says the system covers the roughly 2% of the genome that codes for proteins as well as the much larger non-coding portion involved in regulating gene activity.
The Atlas also introduces an AlphaGenome Variant Impact, or AVI, score. The metric combines predictions from AlphaGenome and AlphaMissense to help researchers prioritise variants for further study.
Read: Google DeepMind Unveils Gemma 4 Open AI Model
Researchers can access the database through a web portal without writing code. The supplied material says access is free for academic and noncommercial research.
Scientific testing and input involved researchers from institutions including the Stowers Institute, Broad Institute and Stanford University.
Read: Google DeepMind’s Genie 3 AI Model Advances Toward AGI with 3D World Creation
Google DeepMind has stressed that the Atlas is a research prioritisation tool rather than a standalone diagnostic system. Its predictions still require laboratory and clinical validation before researchers can draw medical conclusions.
