Open AccessAlphaFold 2: Why It Works and Its Implications for Understanding the Relationships of Protein Sequence, Structure, and Function
Author(s) -
Jeffrey Skolnick,
Mu Gao,
Hongyi Zhou,
Suresh B. Singh
Publication year - 2021
Publication title -
journal of chemical information and modeling
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.24
H-Index - 160
eISSN - 1549-960X
pISSN - 1549-9596
DOI - 10.1021/acs.jcim.1c01114
Subject(s) - computer science , completeness (order theory) , protein data bank (rcsb pdb) , sequence (biology) , computational biology , perspective (graphical) , protein structure , function (biology) , resolution (logic) , domain (mathematical analysis) , artificial intelligence , algorithm , biology , mathematics , genetics , mathematical analysis , biochemistry
AlphaFold 2 (AF2) was the star of CASP14, the last biannual structure prediction experiment. Using novel deep learning, AF2 predicted the structures of many difficult protein targets at or near experimental resolution. Here, we present our perspective of why AF2 works and show that it is a very sophisticated fold recognition algorithm that exploits the completeness of the library of single domain PDB structures. It has also learned local side chain packing rearrangements that enable it to refine proteins to high resolution. The benefits and limitations of its ability to predict the structures of many more proteins at or close to atomic detail are discussed.