DeepMind Releases AI-Predicted Structures for 200 Million Proteins
DeepMind's AlphaFold tool now predicts the structure of nearly every protein known to science, and the company offers this massive database to the public for free. This unprecedented resource promises to accelerate drug discovery and breakthroughs in biology.
DeepMind's AlphaFold tool successfully predicts the structure of nearly every protein known to science, expanding its public database to over 200 million proteins. The Alphabet-owned AI lab offers this massive trove of data to anyone for free, building on its initial 2020 release that stunned the scientific community. By solving one of biology's grand challenges, AlphaFold makes it possible to search for 3D protein structures almost as easily as performing a standard web search.
This latest update provides structural predictions for a vast array of organisms, including plants, bacteria, and animals. DeepMind founder Demis Hassabis highlights that this comprehensive coverage opens up huge opportunities to tackle important global issues such as sustainability, food insecurity, and neglected diseases. Researchers no longer have to spend years experimentally determining these shapes, as the AI instantly provides highly accurate models.
The scientific community eagerly embraces this expanded database as a transformative resource for understanding diseases and accelerating innovation in drug discovery. Scientists already use AlphaFold's previous data to study honeybee health and develop effective malaria vaccines. Experts call AlphaFold the most significant contribution the AI community has ever made to the scientific community, unlocking countless new pathways for biological research.