基于聚类的蛋白质一级结构与三级结构关联研究

An Association Rule Study Between Protein Primary Structure and Tertiary Structure Based on Clustering

为了建立蛋白质一级结构与三维结构间的关联规则,以蛋白质数据库PDB为数据源,采用K均值聚类方法对蛋白质二级结构中α-螺旋、β-折叠和无规则卷曲的疏水值、偶极矩、解离常数3种属性值序列进行分析。结果表明:80%的螺旋和85%的beta折叠都表现出了自身应有的规律,说明聚类分析方法在蛋白质三级结构预测研究中的合理性和有效性。

In order to establish the association rules between protein primary structure and tertiary structure, in this study, using PDB database as the data source, we used K-means clustering method to analyze the values of sequences of hydrophobic rules, dipole moment,dissociation constant which were mapped from α - helix, β - sheet and random coil. The results showed that 80% spiral and 85% beta folding had showed its own rules. This study illustrated that clustering analysis method has rationality and validity in protein structure prediction.