Google DeepMind publishes a predictive map of all 9 billion single-letter DNA changes in the human genome
A petabyte of predictions, free to browse, and explicitly not a diagnosis.
Google DeepMind released AlphaGenome Atlas on Tuesday, a one-petabyte dataset of predicted molecular effects for all 9 billion possible single-letter changes to human DNA, which the lab says is more than 30 times larger than its AlphaFold Database. Each variant carries thousands of gene-regulation predictions across hundreds of human and mouse cell types and tissues, plus a single Variant Impact score that combines AlphaGenome and AlphaMissense outputs. Collaborators who contributed scientific input include the Stowers Institute for Medical Research, the Broad Institute, the University of Exeter, Memorial Sloan Kettering Cancer Center and Boston Children's Hospital. Access is free for non-commercial research through the website now, with commercial access on Google Cloud promised later; DeepMind states the resource is not approved for any clinical use and is not a substitute for medical advice.
Scripture for Reflection
“Great are the works of the LORD; they are pondered by all who delight in them.”
We can slow down and ponder the intricacy of a body we did not design.
