Google DeepMind Unveils Atlas Mapping 9 Billion DNA Variants for Human Genome Research
Updated
Updated · The Verge · Sep 8
Google DeepMind Unveils Atlas Mapping 9 Billion DNA Variants for Human Genome Research
3 articles · Updated · The Verge · Sep 8
Summary
AlphaGenome Atlas launches Tuesday with predictions for every possible single-letter DNA change in the human genome, giving researchers a genome-wide map of how roughly 9 billion variants may affect biology.
The tool extends DeepMind’s AlphaGenome model beyond protein-coding regions into the broader genome, where many stretches regulate gene behavior but are harder to interpret.
Google paired the atlas with a Variant Impact Score to help scientists rank mutations by likely significance and inspect their molecular effects through a web portal, Antigravity and the AlphaGenome interface.
The dataset, built from public human and mouse genome databases, totals about 1 petabyte; DeepMind said precomputing and analyzing variants at that scale took time.
Researchers can access Atlas for noncommercial use starting now, with commercial availability on Google Cloud coming soon as Google pushes AI deeper into biology and drug discovery.
Will DeepMind's massive petabyte-scale genetic atlas revolutionize personalized medicine, or will its blind spots in cellular specificity lead researchers astray?
Can an AI mapping nine billion DNA mutations finally unlock the mysteries of human disease hidden in our noncoding genetic dark matter?