基于序列和局部信息熵的蛋白质折叠速率预测模型  

Prediction Model of the Protein Folding Rate Using Sequence Representation and Local Structural Entropy

作　　者：高建召[1] 胡刚[1] 王奎[1] 沈世镒[1] 

机构地区：[1]南开大学数学科学学院与LPMC,天津300071

出　　处：《工程数学学报》2010年第6期959-966,共8页Chinese Journal of Engineering Mathematics

基　　金：国家自然科学基金(10671100;20836005);刘徽应用数学研究中心;天津大学;南开大学联合研究项目;天津市自然科学基金(07JCZDJC06400)~~

摘　　要：正确预测蛋白质折叠速率对理解蛋白质的折叠机制非常重要。本文从AAindex数据库中的531种残基物理化学性质、序列长度信息和局部结构信息熵中筛选特征,从而提出了一个基于蛋白质序列信息的线性回归模型。针对三种折叠机制two-state,multi-state和mixed-state,用Jackknife验证模型,预测的折叠速率和实验验证的折叠速率相关系数分别为0.790,0.829和0.778。本文结果表明四阶局部结构信息熵和折叠速率有很高的负相关性;蛋白质的长度和蛋白质的折叠速率成反比关系;螺旋的含量会加快蛋白质的折叠过程。对two-state蛋白质β折叠的含量会减慢蛋白质的折叠过程;和其他模型相比,我们提出的线性回归模型具有输入参数少,计算简单,平均绝对误差小的优点。Prediction of protein folding rates is important in understanding the overall folding the mechanism.This article selects the features from 531 physical chemistry properties in the AAindex database,the length of proteins and the local structural entropy,and proposes three sequence-based linear regression models for two-state,multi-state and mixed-state proteins.The correlation between predicted folding rates and experimental folding rates for different folding kinetics is 0.790,0.829 and 0.778,respectively.We show that the tetra-local structural entropy is negatively correlated with the protein folding rate.Length of protein is negatively correlated with the folding rates.Coil content may accelerate the protein folding process and for two-state proteins,the beta-strand content may decelerate the folding process.Compared with other models,our proposed method has advantages in less features,simple computation and smaller mean absolute errors.

关 键 词：蛋白质折叠速率 基因序列的预测方法 局部结构信息熵 线性回归 

分 类 号：O236[理学—运筹学与控制论]