How did DeepMind's AlphaFold predict the structure of 200 million proteins and revolutionize biology?

The answer

DeepMind's AlphaFold 2, which won the CASP14 protein structure prediction competition in December 2020 with a median GDT score of 92.4 out of 100, went on to predict the 3D structures of over 200 million proteins - representing nearly every protein known to science - released as a free public database in July 2022. Experimental methods like X-ray crystallography would have taken an estimated one billion years to achieve the same coverage. The AlphaFold Protein Structure Database has been accessed by over 2 million researchers in 190 countries. The impact on drug discovery has been transformative: pharmaceutical companies use AlphaFold predictions to identify drug targets, design binding molecules, and understand disease mechanisms at a fraction of the traditional cost and time. The breakthrough earned DeepMind co-founder Demis Hassabis and researcher John Jumper the 2024 Nobel Prize in Chemistry, alongside David Baker for computational protein design.