The latest version of DeepMind’s AlphaFold2 algorithm outperformed around 100 other teams in the biennial protein-structure prediction challenge CASP14. The Critical Assessment of protein Structure Prediction (CASP) experiments aim at establishing the current state of the art in protein structure prediction, identifying what progress has been made, and highlighting where future effort may be most productively focused. Protein folding is a major computational challenge with huge potential applications in medicine and health. The results were announced by DeepMind on 30 November.
AlphaFold’s performance marks a turning point for DeepMind, after another program developed by the company owned by Google, AlphaGO defeated the then world No. 1 ranking GO player Ke Jie. DeepMind’s long-term goal is to develop programs capable of achieving broad, human-like intelligence. Tackling grand scientific challenges, such as protein-structure prediction, is one of the most important applications AI can make, DeepMind’s CEO Demis Hassabis says:
I do think it’s the most significant thing we’ve done, in terms of real-world impact.
[…] a new mechanism of cellular organization (video); Demis Hassabis and John Jumper for developing AlphaFold, which accurately predicts the structure of proteins (video); and to Emmanuel Mignot and Masashi […]