bioRxivBioinformaticsPreprint
The computational models of AlphaFold2 and RoseTTAfold carry protein foldability information
Abstract
Protein folding has been a "holy grail" problem of biology for fifty years, and the recent breakthrough from AlphaFold2 and RoseTTAfold set a profound milestone for solving this problem. AlphaFold2 and RoseTTAfold were successful in predicting protein structures from peptide sequences with high accuracy. Meanwhile, although the protein folding problem also cares about the kinetic pathways of protein folding, AlphaFold2 and RoseTTAfold were not trained with this functionality. Considering that their training sets contain proteins from foldable sequences and sequence evolutionary information, we wondered if the computational models from AlphaFold2 and RoseTTAfold might carry protein foldability information. To test this idea, we systematically predicted the structural models of 149 circular permutants and 148 alanine insertion mutants of the 149-residue dihydrofolate reductase of Escherichia coli with AlphaFold2 and RoseTTAfold. Our data showed that although AlphaFold2 and RoseTTAfold could not directly identify unfoldable proteins, the structural variations of computational models are correlated with protein foldability. Furthermore, this correlation is independent of secondary structures. Most importantly, the structural variations of computational models are quantitatively correlated with protein foldability but not protein function. Our work could be of great value to the design of circular permutants, the design of fragment complementary proteins, the design of novel proteins, and the development of computational tools for predicting protein folding kinetics. HighlightsO_LIAlphaFold2 and RoseTTAfold could not directly identify unfoldable proteins. C_LIO_LIThe structural variations of computational models are correlated with protein foldability. C_LIO_LIThis correlation is independent of secondary structures. C_LIO_LIThe structural variations of computational models are quantitatively correlated with protein foldability but not protein function. C_LI
- organism: E. coli
- assay: protein structure
- research focus: protein structure prediction
- research focus: benchmarking
- approach: deep learning
- approach: statistical model
- Citations
- 4
Reproducibility ledger
Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.
Sign in to file a reproducibility report.
Related papers
- AlphaFold2's training set powers its predictions of fold-switched conformations
Joseph W. Schafer, Lauren L. Porter·bioRxiv·Oct 11, 2024
- Efficient and accurate prediction of protein structure using RoseTTAFold2
M. Baek, I. Anishchenko, … F. DiMaio·bioRxiv·May 24, 2023
- GDFold2: a fast and parallelizable protein folding environment with freely defined objective functions
Tianyu Mi, Nan Xiao, Haipeng Gong·Protein SciencePeer-reviewed·Jan 28, 2025
- Applications of AlphaFold beyond Protein Structure Prediction
Yuan Zhang, Peizhao Li, … Jinfeng Zhang·bioRxiv·Nov 3, 2021
- AlphaFold2 knows some protein folding principles
Liwei Chang, Alberto Pérez·Nature CommunicationsPeer-reviewed·Dec 1, 2025
- The impact of AlphaFold on experimental structure solution
Maximilian Edich, David C. Briggs, … Andrea Thorn·bioRxiv·Apr 7, 2022