Deeper collaboration between EMBL-EBI and Google DeepMind brings updates to the AlphaFold Database
EMBL’s European Bioinformatics Institute (EMBL-EBI) and Google DeepMind today announced the renewal of their collaboration on the AlphaFold Protein Structure Database (AFDB) – a landmark partnership that continues to deliver high-quality protein structure predictions to the global scientific community.
To mark this continued partnership, the AlphaFold Database has launched a significant update, which synchronises it with the latest protein sequence data in UniProt, one of over 40 data resources managed and hosted by EMBL-EBI.
Deepening collaboration to advance protein science
The AlphaFold Database , launched in 2021, is a collaboration between Google DeepMind and EMBL-EBI. It provides open access for the global scientific community to highly accurate protein structure predictions from the Nobel-prize-winning AlphaFold AI system.
The AlphaFold Database contains protein structure predictions for over 200 million proteins, and has been used by over three million people in over 190 countries. It helps scientists understand protein structures and narrows the gap between the vast number of known protein sequences and the much smaller number of experimentally determined structures.
“The AlphaFold Database has been a catalyst for protein research,” said Jo McEntyre, EMBL-EBI’s Interim Director. “By renewing our partnership with Google DeepMind, we can continue to make protein structure predictions openly available, accessible and useful to all. This collaboration exemplifies how academia and industry can work together to advance scientific discovery.”
“The partnership between Google DeepMind and EMBL-EBI is a powerful example of how, by working with partners, AI innovation can be responsibly scaled to have the greatest scientific impact,” said Anna Koivuniemi, Head of Google DeepMind Impact Accelerator. “Renewing this collaboration ensures that the global research community will continue to have open access to state-of-the-art protein structure predictions. We are incredibly proud to continue this work and support the discoveries it will unlock.”
New data and functionalities
To mark the partnership renewal, the AlphaFold Database (AFDB) has been synchronised with the UniProt protein sequence Knowledgebase (release 2025_03). UniProt is a longstanding collaboration between EMBL-EBI, the Swiss Institute of Bioinformatics (SIB), and the Protein Information Resource (PIR).
This AFDB update ensures AlphaFold coverage for newly discovered proteins and incorporates protein isoforms, providing a more complete view of proteome diversity. This is the first time isoforms are available in AFDB, enabling users to explore functionally-distinct variants of the same protein and gain deeper insights into tissue-specific roles, regulatory mechanisms and disease-associated differences across the proteome.
In addition, the multiple sequence alignments (MSAs) used to generate each prediction are now available for download, offering a rich new layer of data for evolutionary and structural analyses.