基于CGR的蛋白质相似性比较  

Protein Similarity Comparison Based on CGR

作　　者：徐占[1] 董洪伟[1] 

机构地区：[1]江南大学信息工程学院,无锡214122

出　　处：《计算机工程》2010年第14期233-234,237,共3页Computer Engineering

摘　　要：从蛋白质结构特性出发,利用结构字母表和CGR游走技术将蛋白质三维结构信息转换到二维坐标空间中。通过分析所得图像找出蛋白质分子的主体结构,获得各结构点在CGR图中的坐标,利用Hausdorff距离判定要比较的蛋白质对象相似性。该方法实现了蛋白质相似性比较的结构-序列模式转变,利用Hausdorff距离比较两点集间相似性的优势,为蛋白质相似性比较提供了一种简便有效的方法。Starting from the characteristics of protein structures,a new method based on structure alphabet and Chaos Game Representation（CGR） is provided.Its can find the main structure of the protein easily by CGR,and make use of CGR and Hausdorff distance to complete protein similarity comparison.Compared with other methods based on amino acid sequence,this method can resolve the problem that different proteins are classified into one family just for the similar sequence of amino acid.The proteins with similar sequence of amino acid may be belong to different protein families,because they have different structures.The functions of protein depend on their structures,so this method can get better result of similarity compared with that of amino acid sequence.

关 键 词：结构字母表 主体结构 HAUSDORFF距离 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]