With the breakthrough of AlphaFold2 and the publication of AlphaFold DB, the protein structure prediction has made remarkable progress. However, AlphaFold2 models tend to represent only a single static structure, and accurately predicting multiple conformations remains a challenge.
In this work, we proposed a multiple conformational states folding method using the distance-based multi-objective evolutionary algorithm framework, named MultiSFold.
On a developed benchmark testset containing 81 proteins with two representative conformational states, the success ratio of MultiSFold predicting multiple conformations was 70.4% while that of AlphaFold2 was 9.88%, which may indicate that conformational sampling combined with knowledge gained through deep learning has the potential to produce conformations spanned the range between two experimental structures.
In addition, MultiSFold was tested on a single static structure benchmark testset of 244 human proteins and participated in a blind test of CAMEO to verify our hypothesis that multiple competing optimization objectives may be required in the conformation search to improve the prediction accuracy of a single static structure.
MultiSFold On-line Server
Ming-Hua Hou, Chun-Xiang Peng, Xiao-Gen Zhou, Biao Zhang and Gui-Jun Zhang*.
Multi contact-based folding method for de novo protein structure prediction.
Briefings in Bioinformatics, 2022, 23(1): bbab463.
Ming-Hua Hou, Si-Rong Jin, Xin-Yue Cui, Chun-Xiang Peng, Kai-Long Zhao, Le Song* and Gui-Jun Zhang*.
Protein multiple conformations prediction using multi-objective evolution algorithm.