Story perspectives

AI Boosts AlphaFold Database with 1.7M Protein Predictions

3/19/2026

Summary

Story summary

Google DeepMind, EMBL-EBI, NVIDIA, and Seoul National University have collaborated to enhance the AlphaFold Database with AI-predicted protein complex structures.
The update adds 1.7 million high-confidence homodimer predictions, focusing on proteins crucial for human health.
The database now contains about 200 million predictions, aiding researchers in understanding protein interactions and functions.
The effort aims to democratize access to protein complex data and empower scientists worldwide to advance health research.

1 of 1

Story summary

Google DeepMind, EMBL-EBI, NVIDIA, and Seoul National University have collaborated to enhance the AlphaFold Database with AI-predicted protein complex structures.
The update adds 1.7 million high-confidence homodimer predictions, focusing on proteins crucial for human health.
The database now contains about 200 million predictions, aiding researchers in understanding protein interactions and functions.
The effort aims to democratize access to protein complex data and empower scientists worldwide to advance health research.