CAJ | 학술논문

关于二硫键是由蛋白质的两个半胱氨酸之间配对形成的一种共价键,可以存在于同一条蛋白质多肽键内,也可以存在于不同的多肽键之间.二硫键的形成是蛋白质折叠过程中的重要步骤,同时影响蛋白质折叠的速率和途径.因此利用计算机方法预测二硫键连接模式有非常重要的意义.采用一种新的方法预测二硫键连接模式.结合序列多重特征向量和通过PSIPRED得出的二级结构预测信息.由于原方法会产生高维数据,使用PCA进行降维,在降维后的低维数据上采用支持向量回归技术(SVR)来预测二硫键连接模式.结果显示,上述方法提高了预测精度.
관우이류건시유단백질적량개반광안산지간배대형성적일충공개건,가이존재우동일조단백질다태건내,야가이존재우불동적다태건지간.이류건적형성시단백질절첩과정중적중요보취,동시영향단백질절첩적속솔화도경.인차이용계산궤방법예측이류건련접모식유비상중요적의의.채용일충신적방법예측이류건련접모식.결합서렬다중특정향량화통과PSIPRED득출적이급결구예측신식.유우원방법회산생고유수거,사용PCA진행강유,재강유후적저유수거상채용지지향량회귀기술(SVR)래예측이류건련접모식.결과현시,상술방법제고료예측정도.
Disulfide bonds are primary covalent crosslinks between two cysteine residues in proteins,it can occur in a protein peptide bond or between different protein peptide bonds.To many proteins,disulfide connectivity is the character of the final folding protein structure.It's an important step for the folding of proteins and it also affects the folding rate and way.Therefore,there is a great need to develop computational methods capable of accurately predicting disulfide connectivity patterns in proteins that have potentially important applications.A novel method is used to predict disulfide connectivity patterns from protein primary sequences,and a support vector regression(SVR)approach is used based on multiple sequence feature vectors to predict secondary structure information by the PSIPRED program.Since above method generates too many higher dimensional data,PCA is used to reduce feature dimensions.Compared to the former method without feature reduction,the accuracy is improved by using the method.