多特征框架下的蛋白质相似性比较与分类  

Protein Similarity Comparison and Classification Based on Multiple Characters Framework

作　　者：徐占[1] 董洪伟[1] 

机构地区：[1]江南大学信息工程学院,江苏无锡214122

出　　处：《工程图学学报》2010年第1期185-190,共6页Journal of Engineering Graphics

摘　　要：蛋白质的相似性比较是研究蛋白质结构和功能的重要手段。与以往基于构建距离矩阵来研究蛋白质空间结构相似性的方法不同,文章从蛋白质的基本组成单元——氨基酸入手,通过对氨基酸以及蛋白质多肽链中的特殊结构的分析,综合考虑骨架原子Cα数、突变原子数、亲水微粒数、螺旋数四个方面,依托模糊数学等价矩阵理论,运用传递闭包的方法,提出了一种全新的相似性比较与分类方法。Protein similarity comparison is an important way to study protein structure and function. Many methods are proposed which are capable of comparing all available protein structure. Being different from those methods, a new method based on multiple characters is provided. There are four characteristics and some internal relationships among the four characteristics. So a method based on fuzzy equivalent matrix is proposed in which the four characteristics are considered comprehensively.

关 键 词：计算机应用 相似性比较 多特征框架 模糊等价矩阵 

分 类 号：Q51[生物学—生物化学]