- Course overview
- Search within this course
- Understanding the structure prediction page
- Citing the database
- Getting help and providing feedback
- Your feedback
All materials are free cultural works licensed under a Creative Commons Attribution 4.0 International (CC BY 4.0) license, except where further licensing details are provided.
What is the AlphaFold Database
In 2020, DeepMind’s AlphaFold 2 system marked a monumental breakthrough by utilising deep learning to predict protein structures with remarkable accuracy (Jumper, J., et al, 2021). This paradigm shift allowed researchers to gain structural insights in minutes rather than months or years.
To make these predictions freely and openly available to the global scientific community, DeepMind and EMBL-EBI collaborated to create the AlphaFold Protein Structure Database (AFDB).
As of 2025, the database hosts over 241 million high-accuracy protein structure predictions, covering the vast majority of the known protein universe.
See the homepage of the AlphaFold Protein Structure Database below and click on the 
to find useful information.
The AlphaFold Protein Structure Database homepage.
The AFDB contains several distinct collections of protein structure predictions. The dataset includes predictions for protein structures from over 1 million organisms, including model organisms and pathogens of interest from the World Health Organisation (WHO).