DeepMind Releases AlphaFold Predictions for Protein Structure
In July 2022, DeepMind released the AlphaFold protein structure predictions for almost all known proteins. This is a significant contribution to advancing scientific knowledge and our understanding of biology.
Proteins are like tiny biological machines and understanding their structure helps us understand their function. By sharing predictions for the shape of every protein in the human body and the proteins of 20 other organisms, we are doubling humanity’s understanding of the human proteome and revealing new insights into the structure of proteins in other important organisms.
This development is considered the most significant contribution AI has made to advancing scientific knowledge and will support future advances in biology and medicine. AlphaFold has demonstrated the capability to accurately predict the shape of proteins at scale and in minutes, providing a valuable resource for researchers.
The AlphaFold team has worked tirelessly for five years, culminating in the release of this groundbreaking technology to the public. This new protein almanac is now available to scientists worldwide and has already shown promise in various research areas, including finding cures for diseases, studying antibiotic resistance, and addressing climate change.
This work represents the culmination of tremendous effort and is a testament to the potential of AI to revolutionize scientific research in the 21st century. To learn more or explore AlphaFold, visit Nature for our peer-reviewed papers and open-source code. The AlphaFold Protein Structure Database is hosted by EMBL-EBI, offering a valuable resource for scientists worldwide. We are excited to hear how AlphaFold has been useful in your research and welcome your feedback at alphafold@deepmind.com.