DeepMind and EMBL Unveil Open Database of Human Protein Structures
DeepMind and EMBL partner to release the most complete database of predicted 3D human protein structures. This open-source tool more than doubles the known high-accuracy protein structures to accelerate global scientific research.
DeepMind and the European Molecular Biology Laboratory (EMBL) partner to provide the scientific community with the most complete and accurate database of predicted human protein structures to date. This open database covers all approximately 20,000 proteins expressed by the human genome, giving researchers free and open access to a massive trove of biological data. By revealing the three-dimensional shapes of these essential molecules, the project unlocks powerful new tools for structural biologists.
The AlphaFold Protein Structure Database dramatically expands the accumulated knowledge of human protein structures, easily more than doubling the number of high-accuracy models available to the global scientific community. This achievement builds upon AlphaFold's previous success in solving the 50-year-old grand challenge of protein structure prediction. The database stands on the shoulders of generations of scientific work, ranging from early protein imaging to modern artificial intelligence innovations.
Early partners already utilize these predicted structures to tackle significant global challenges, including researching neglected diseases, studying antibiotic resistance, and developing methods for recycling single-use plastics. Because proteins act as the fundamental building blocks of life that underpin every biological process, understanding their exact shapes accelerates research across a huge variety of scientific fields. This comprehensive release ultimately heralds a new era for AI-enabled biology.