Google DeepMind releases AlphaGenome Atlas, mapping molecular effects of 9 billion human DNA variants

Google DeepMind introduced AlphaGenome Atlas, a platform that predicts the molecular effects of every possible single-letter DNA variant in the human genome—approximately 9 billion variants. The atlas is built by scaling AlphaGenome’s predictions across the entire genome, enabling researchers to explore how genetic variations impact biology at a molecular level.

The atlas contains thousands of molecular effect predictions for each variant, covering multiple aspects of gene regulation across hundreds of human and mouse cell types and tissues. This breadth aims to provide a holistic view of how single-letter changes can disrupt or alter biological processes.

To help researchers quickly prioritize variants, DeepMind released the AlphaGenome Variant Impact (AVI) score, which combines predictions from AlphaGenome and AlphaMissense into a single metric for ranking and interpreting variant effects.

AlphaGenome Atlas is accessible today through a free website portal, the AlphaGenome API, and as a skill in Google Antigravity, designed to be intuitive for researchers regardless of coding experience.

The platform builds on the model’s existing capabilities, which have already seen widespread use in research, but expands its reach by precomputing predictions at scale. DeepMind notes that trusted external collaborators have used the atlas to identify and experimentally verify key variants in unsolved rare disease research and to find rare variants linked to common traits.

The atlas is a 1-petabyte dataset—more than 30 times larger than the AlphaFold Database at its 2022 expansion—which grew from roughly 190,000 experimental structures to over 200 million protein structure predictions. Like the AlphaFold Database, AlphaGenome Atlas provides visualizations and tools intended to make large-scale analysis accessible to non-coding researchers.