For the first time, a biologist in Africa, South Asia, or Latin America can access for free the same genomic-prediction tools as a researcher at MIT. Google DeepMind released on September 8, 2026 AlphaGenome Atlas, an open database aimed at reducing global scientific inequalities. The platform contains predictions of the effects of 9 billion possible genetic mutations on human DNA.
Genomics, a field once reserved for wealthy Western laboratories
Decoding the human genome has required substantial resources until now. Since the completion of the Human Genome Project in 2003, scientists have had the full sequence of our DNA, but not its functional “reading.” Pushmeet Kohli, vice president of research at Google DeepMind, summarizes: “We bought the book, but we don’t know how to read it.”
Analyzing genetic variants required costly sequencing equipment, advanced bioinformatics expertise, and years of bench work. Only research centers with large budgets could systematically explore the non-coding regions of the genome, which account for 98% of the human genome. An isolated researcher in Uganda or Bangladesh had no way to access these analytical capabilities, deepening global scientific inequality.
AlphaGenome Atlas: a portal accessible to everyone
AlphaGenome Atlas radically changes the game. The platform freely provides predictions of the effects of all possible single-letter substitutions in the human genome. Each genetic variant includes on average 27,000 individual predictions spanning gene expression and DNA transcription across hundreds of cell types and tissues, occupying 1 petabyte of data — more than 30 times the size of AlphaFold.
The major innovation lies in accessibility. Unlike traditional bioinformatics tools, AlphaGenome Atlas operates through a simple web portal that does not require programming skills. A clinical biologist in Lagos, Dhaka, or Lima can query the database, download the relevant predictions, and quickly identify suspect genetic variants. Dr. Gareth Hawkes of the University of Exeter notes: “The human genome represents an enormous research space. We can use it to shrink the haystack.” His team used Atlas to analyze more than 54,000 participants from the UK Biobank, uncovering an additional 22% of non-coding genetic associations.
Free access for universities and researchers around the world
Google DeepMind guarantees free access for all academic researchers, embracing an open philosophy similar to the Human Genome Project. The goal: accelerate global scientific discovery by enabling every team to contribute. For example, researchers from the Broad Institute used Atlas to reclassify a variant in the DNM1 gene as likely pathogenic in a case of epileptic encephalopathy.
The availability of AlphaGenome Atlas raises the question of whether genomic knowledge should be a common good or a private asset. The tool introduces the AlphaGenome Variant Impact (AVI) score, standardizing predictions for coding and non-coding regions, which facilitates international comparisons and could harmonize genetic-diagnostic protocols.
Ewan Birney, director of the European Bioinformatics Institute (EMBL), stresses the importance of openness: “These tools reach their full value when they are open and connected to the broadest possible data ecosystem.” Google DeepMind, by making Atlas accessible with no technical barriers, encourages a global, collaborative scientific community.
