DeepMind Releases Complete 3D Database of Human Proteins
DeepMind and its partners launch a free database featuring the 3D structures of nearly every human protein, predicted by its AlphaFold AI. This massive release mirrors the historic human genome project and promises to accelerate biological and medical research worldwide.
DeepMind and its research partners release a groundbreaking database containing the 3D structures of nearly every protein in the human body. This massive repository relies on predictions generated by AlphaFold, the revolutionary artificial intelligence system that solves the complex protein folding problem. DeepMind founder Demis Hassabis calls this the most significant contribution AI has ever made to advancing scientific knowledge.
The AlphaFold Protein Structure Database acts as a freely available resource for scientists across countless disciplines, much like the historic human genome project from the early 2000s. While sequencing the genome provides the basic blueprint of human life, understanding the proteome reveals how those genetic instructions actually build physical molecular machines. Proteins consist of 20 amino acids that fold into intricate, functional shapes rather than existing as simple linear code.
Because this complex folding process determines exactly how proteins function within our bodies, predicting these 3D structures opens up entirely new pathways for medical research and drug development. The current database includes hundreds of thousands of protein sequences, but the collaborative team plans to expand it into a comprehensive protein almanac containing millions of entries. Researchers around the world immediately access this data to accelerate their work in understanding and treating countless diseases.