Science
Featured

Google DeepMind

AlphaFold: Solving the 50-Year-Old Protein Folding Problem

200M+ protein structures predicted
95%+ accuracy vs. experimental methods
Freely available to all researchers
1M+ researchers using the database
DeepMind team awarded Nobel Prize in Chemistry (2024)

The Problem

Determining a protein's 3D structure from its amino acid sequence (the 'protein folding problem') was a 50-year unsolved challenge in biology. Experimental structure determination takes months and costs $100,000+. With millions of known proteins, experimental methods couldn't scale.

The Solution

DeepMind's AlphaFold 2 uses a deep learning architecture that processes amino acid sequences and predicts atomic-level 3D structures with near-experimental accuracy. AlphaFold 3 (2024) expanded to predict structures of DNA, RNA, and protein complexes with small molecules.

The Outcome

AlphaFold 2 predicted structures for 200M+ proteins — essentially the entire known protein universe. Made freely available to researchers worldwide. Accelerated drug discovery, enzyme engineering, and fundamental biology research.

Key Metrics

  • 200M+ protein structures predicted
  • 95%+ accuracy vs. experimental methods
  • Freely available to all researchers
  • 1M+ researchers using the database
  • DeepMind team awarded Nobel Prize in Chemistry (2024)
ScienceDrug DiscoveryBiologyDeep LearningResearch