Prediction And Application Analysis Of Protein Folding Rate

It is of great significance to give a panoramic view of the relationship between protein evolution,function and folding rate on the genome scale,which is of great significance to understand the central issue of protein science of"sequence-structure-function".However,limited by the development of experimental measurement technology,the experimental data of folding rate of proteome can’t be produced at present,which makes us still know little about the relationship between protein folding rate and species and subcellular location.Therefore,it is a feasible strategy to build a high-precision folding rate prediction model first,and then discuss it based on the prediction results of the model.Based on this strategy,the following work has carried out in this paper.Firstly,it is assumed that the folding element of protein is a natural secondary structure,and the organization process of the folding element is folding.An empirical parameter,coupling of secondary structure number and nonlocal interaction(CSNI),was constructed to characterize the height of the folding free energy barrier.Furthermore,a prediction model of folding rate was constructed according to the transition state theory.On the existing experimental data set,the goodness of fit of CSNI model for folding rate is R~2=0.73,and the correlation is better than the existing typical models.Secondly,based on the CSNI model,the folding rate of proteome of nearly 30 species including archaea,prokaryotes and eukaryotes was estimated,and the folding time scales of protein group of different taxonomies and different species under the same classification were roughly given.At the same time,we also analyzed the folding rate of E.coli and yeast subcellular localization protein.The results showed that:1)from the taxonomic point of view,the average folding rate of proteome was as follows:Archaea<prokaryotes<eukaryotes.This result is consistent with the conclusion pointed out by Debes et al.that the folding rate of protein increases with evolution.2)According to the type of folding dynamics,proteins were divided into two-state and multi-state folding.According to 159 experimental data,the folding rate of two-state protein in prokaryote was slightly higher than that in eukaryote,but the difference was not significant.The folding rate of polymorphic protein in eukaryote was significantly higher than that in prokaryote.However,the CSNI model estimates that the order of 1)is maintained in both two-state and multi-state.This may be related to the poor prediction accuracy of CSNI model for the folding rate of two-state protein.3)There were significant differences in the average protein folding rates among four groups of four subcellular positions in E.coli,but no significant differences among five subcellular positions in yeast.This may be due to the multi localization of most proteins in yeast.To sum up,this paper constructed an accurate prediction model of folding rate on the experimental data set,and estimated the distribution of folding rate of proteome in many species by using this model.The conclusions provide some useful information for studying the relationship between protein folding and evolution.