基于复杂网络理论的PPI网络拓扑分析

蛋白质相互作用在生命活动中起核心作用，由蛋白质相互作用构成的PPI网络的拓扑特性分析是后基因组时代最重要的研究课题之一。应用复杂网络理论对DIP数据库中7个物种的8个PPI网络的拓扑结构进行分析与研究。分析结果表明，这些PPI网络具有较小的平均路径长度和较高的聚集系数，其度分布服从幂规律，即p（k）=ak^-r，其中r大于1小于3，a近似等于1±0.5，表现出典型的无标度性，并具有高的异质性。其中平均度大于3.5的5个PPI网络对随机删除不超过10%的顶点都具有很好的鲁棒性，但对有选择的删除2%的高度顶点就开始表现出极弱的抗攻击性。

Analysis of PPI network topology based on complex network theory

Protein-protein interactions play an important role in life.Analysis of the topology characters of PPI network composed of protein-protein interactions is one of the most important issues in the post-genomic era.This paper analyzes the topologies of 8 PPI networks of 7 species in DIP database by using complex network theory.Analysis shows that these PPI networks have small average path length and large clustering coefficients.The degree distributions of these networks follow a power law as given by p（k）=ak^-r（1<r<3，a≈1±0.5）.All these PPI networks are scale-free and idiosyncratic.All the PPI networks with average degree larger than 3.5 are robust against the random removal of not more than 10% nodes，but are vulnerable to the removal of 2% nodes of high degree.