DeepMind AI Solves 50-Year-Old Protein Folding Challenge
Alphabet-owned DeepMind develops an AI system called AlphaFold that accurately predicts protein structures in days, potentially revolutionizing drug discovery and disease understanding.
Alphabet-owned DeepMind develops an artificial intelligence system called AlphaFold that accurately predicts the 3D structures of proteins in a matter of days. This breakthrough solves a 50-year-old grand challenge in computer science and biology, achieving unparalleled levels of accuracy compared to experimental gold standards. The AI system enters a competition organized by CASP, a group that monitors progress in protein structure prediction.
Predicting how proteins fold is crucial because a protein's shape determines its function, and nearly all diseases relate to how proteins operate in the body. By solving this complex biological puzzle, AlphaFold opens new doors for understanding conditions like cancer and dementia. Experts note that this advancement holds major implications for the future of drug design and the emerging field of protein design.
DeepMind, which employs around 1,000 staff and operates with next to no revenue, stands as a leader in the global AI race alongside companies like Facebook AI Research and OpenAI. The achievement earns praise from Google CEO Sundar Pichai and highlights the potential of AI to tackle fundamental scientific problems. DeepMind co-founder Demis Hassabis emphasizes that this milestone aligns with the ultimate vision behind the research lab.