Protein structure prediction by AlphaFold2: are attention and symmetries all you need?. Issue 8 (4th August 2021)

• Record Type: Journal Article
• Authors: Bouatta, Nazim; Sorger, Peter; AlQuraishi, Mohammed
• Is Part Of: Acta crystallographica. Volume 77:Issue 8(2021)
• Page Start: 982
• Page End: 991
• Publication Date: 2021-08-04
• Subjects: AlphaFold2 -- protein structure prediction -- CASP14; X-ray crystallography -- Periodicals; Crystallography -- Periodicals; Molecular biology -- Periodicals; Molecular structure -- Periodicals; Biomolecules -- Structure -- Periodicals; Cytology -- Periodicals; Biomolecules -- Structure; Crystallography; Cytology; Molecular biology; Molecular structure; X-ray crystallography; Periodicals; 548
• DOI: 10.1107/S2059798321007531
• Languages: English
• ISSNs: 2059-7983

Abstract

Abstract : This review discusses the AlphaFold 2 system for protein structure prediction, including its conceptual and methodological advances, its amenability to interpretation and its achievements in the last Critical Assessment of protein Structure Prediction (CASP14) experiment. Abstract : The functions of most proteins result from their 3D structures, but determining their structures experimentally remains a challenge, despite steady advances in crystallography, NMR and single‐particle cryoEM. Computationally predicting the structure of a protein from its primary sequence has long been a grand challenge in bioinformatics, intimately connected with understanding protein chemistry and dynamics. Recent advances in deep learning, combined with the availability of genomic data for inferring co‐evolutionary patterns, provide a new approach to protein structure prediction that is complementary to longstanding physics‐based approaches. The outstanding performance of AlphaFold 2 in the recent Critical Assessment of protein Structure Prediction (CASP14) experiment demonstrates the remarkable power of deep learning in structure prediction. In this perspective, we focus on the key features of AlphaFold 2, including its use of (i) attention mechanisms and Transformers to capture long‐range dependencies, (ii) symmetry principles to facilitate reasoning over protein structures in three dimensions and (iii) end‐to‐end differentiability as a unifying framework for learning from protein data. The rules of protein folding are ultimately encoded in the physical principles that underpin it; to conclude, the implications of having a powerful computational model for structure prediction that does not explicitly rely on those principles are discussed.