| 一种新的蛋白质结构类预测方法 |
| |
| 引用本文： | 李楠,李春.一种新的蛋白质结构类预测方法[J].生物信息学,2012,10(4):238-240. |
| |
| 作者姓名： | 李楠 李春 |
| |
| 作者单位： | 渤海大学数理学院,辽宁锦州,121000 |
| |
| 基金项目： | 辽宁省高等学校杰出青年学者成长计划 |
| |
| 摘 要： | 基于氨基酸的16种分类模型,给出蛋白质序列的派生序列,进而结合加权拟熵和LZ复杂度构造出34维特征向量来表示蛋白质序列。借助于贝叶斯分类器对同源性不超过25%的640数据集进行蛋白质结构类预测,准确度达到71.28%。
|
| 关 键 词： | 蛋白质结构类预测 氨基酸 加权拟熵 LZ复杂度 贝叶斯分类器 |
A new method for predicting protein structural classes |
| |
| LI Nan
,
LI Chun.A new method for predicting protein structural classes[J].China Journal of Bioinformation,2012,10(4):238-240. |
| |
| Authors: | LI Nan LI Chun |
| |
| Institution: | (College of Mathematics and Physics, Bohai University , Jinzhou , Liaoning 121000 ) |
| |
| Abstract: | Based on 16 kinds of classifications of the amino acids, we obtain the derived sequences of a protein se- quence. Combining the weighted pseudo entropy with Lempel - Ziv complexity, we construct a 34 - D feature vec- tor to represent a protein sequence. Then it is applied to predict the protein structural classes by mens of the Bayes classifier. The dataset includes 640 sequences that share sequence identity below 25%. The accuracy is 71.28%. |
| |
| Keywords: | Prediction of protein structural classes Amino acid The weighted pseudo entropy LZ complexity Bayes classifier |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
