| 基于添加模体信息和功率谱密度的组合向量预测27类蛋白质折叠子 |
| |
| 引用本文： | 刘雷,胡秀珍.基于添加模体信息和功率谱密度的组合向量预测27类蛋白质折叠子[J].生物物理学报,2010(9). |
| |
| 作者姓名： | 刘雷 胡秀珍 |
| |
| 作者单位： | 内蒙古工业大学理学院; |
| |
| 基金项目： | 国家自然科学基金项目(30960090); 内蒙古自治区高等学校科学研究项目(NJZY08059)~~ |
| |
| 摘 要： | 以序列相似性低于40%的1895条蛋白质序列构建涵盖27个折叠类型的蛋白质折叠子数据库,从蛋白质序列出发,用模体频数值、低频功率谱密度值、氨基酸组分、预测的二级结构信息和自相关函数值构成组合向量表示蛋白质序列信息,采用支持向量机算法,基于整体分类策略,对27类蛋白质折叠子的折叠类型进行预测,独立检验的预测精度达到了66.67%。同时,以同样的特征参数和算法对27类折叠子的4个结构类型进行了预测,独立检验的预测精度达到了89.24%。将同样的方法用于前人使用过的27类折叠子数据库,得到了好于前人的预测结果。
|
| 关 键 词： | 模体频数 功率谱密度 支持向量机 蛋白质折叠子 蛋白质结构类型 |
Predicting 27-Class Protein Folds Based on Combined Vector of Adding Motif and Power Spectral Density |
| |
| LIU Lei,HU Xiuzhen College of Sciences,Inner Mongolia University of Technology,Hohhot.Predicting 27-Class Protein Folds Based on Combined Vector of Adding Motif and Power Spectral Density[J].Acta Biophysica Sinica,2010(9). |
| |
| Authors: | LIU Lei HU Xiuzhen College of Sciences Inner Mongolia University of Technology Hohhot |
| |
| Institution: | LIU Lei,HU Xiuzhen College of Sciences,Inner Mongolia University of Technology,Hohhot 010051 |
| |
| Abstract: | A New protein fold dataset containing 1895 proteins with sequence identity below 40%,and classified into 27 fold types,was built for prediction.Based on protein sequence,by using motif frequency,low frequency of power spectral density,amino acid composition,predicted secondary structure and values of auto-correlation function as combined vector,an approach of support vector machine for predicting 27-class protein folds based on ensemble classifier is proposed.Overall predicting accuracy reaches 66.67% in in... |
| |
| Keywords: | Motif frequency Low frequency of power spectral density Support vector machine Protein fold Protein structure class |
| 本文献已被 CNKI 等数据库收录！ |
