John Jumper, Richard Evans, Alexander Pritzel, Tim Green, Michael Figurnov, Olaf Ronneberger, Kathryn Tunyasuvunakool, Russ Bates, Augustin Žídek, Anna Potapenko, Alex Bridgland, ClemensMeyer, SimonA.A.Kohl, Andrew J.Ballard, Andrew Cowie, BernardinoRomera-Paredes, Stanislav Nikolov, Rishub Jain, Jonas Adler, Trevor Back, Stig Petersen, David Reiman, Ellen Clancy, Michal Zielinski, Martin Steinegger, Michalina Pacholska, Tamas Berghammer, Sebastian Bodenstein, David Silver, Oriol Vinyals, Andrew W. Senior, Koray Kavukcuoglu, Pushmeet Kohli & Demis Hassabis

NATURE Published online 15 July 2021 (Nature | www.nature.com)

Abstract

Proteins are essential to life, and understanding their structure can facilitate a mechanistic understanding of their function. Through an enormous experimental efort1–4 , the structures of around 100,000 unique proteins have been determined5 , but this represents a small fraction of the billions of known protein sequences6,7 . Structural coverage is bottlenecked by the months to years of painstaking efort required to determine a single protein structure. Accurate computational approaches are needed to address this gap and to enable large-scale structural bioinformatics. Predicting the 3-D structure that a protein will adopt based solely on its amino acid sequence, the structure prediction component of the ‘protein folding problem’8 , has been an important open research problem for more than 50 years9 . Despite recent progress10–14, existing methods fall far short of atomic accuracy, especially when no homologous structure is available. Here we provide the frst computational method that can regularly predict protein structures with atomic accuracy even where no similar structure is known. We validated an entirely redesigned version of our neural network-based model, AlphaFold, in the challenging 14th Critical Assessment of protein Structure Prediction (CASP14)15, demonstrating accuracy competitive with experiment in a majority of cases and greatly outperforming other methods. Underpinning the latest version of AlphaFold is a novel machine learning approach that incorporates physical and biological knowledge about protein structure, leveraging multi-sequence alignments, into the design of the deep learning algorithm.

See www.nature.com

Figure 1: AlphaFold produces highly accurate structures. (a) AlphaFold’s performance on the CASP14 set (N=87 protein domains) relative to the top-15 entries (out of 146), group numbers correspond to the numbers assigned to entrants by CASP; error bars represent the 95% confidence interval of the median, estimated with 10,000 bootstrap samples. (b) Our prediction of CASP14 target T1049 (blue) compared to the true (experimental) structure (green). Four residues from the C-terminus of the crystal structure are B-factor outliers and are not depicted. (c) An example of a well predicted zinc binding site (AlphaFold has accurate side chains even though it does not explicitly predict the zinc ion). (d) CASP target T1044, a 2,180-residue single chain, predicted with correct domain packing (prediction made after CASP using AlphaFold without intervention). (e) Model architecture. Arrows show the information flow among the various components described in this paper. Array shapes are shown in brackets with s: number of sequences, r: number of residues and c: number of channels.