Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain
the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in
Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles
and JavaScript.
AlphaFold and RoseTTAFold have delivered a revolutionary advance for protein structure predictions, but the implications for drug discovery are more incremental. For now.
The email came out of the blue, titled only “Zoom invite”. But DeepMind’s John Jumper breathed a sigh of relief on opening it. After months of stressing about CASP14 — a biennial competition to predict the structure of proteins on the basis of only their amino acid sequences — the results were clear. His group had “performed amazingly well … both relative to other groups and in absolute model accuracy,” the conference organizers had written.