网络对称性及其在蛋白质交互网络中的应用.pdfVIP

iii 摘 要 后基因组时代，基因序列已经被破译，但生物功能的奥秘还没有解开。随着 生物实验技术的快速发展，产生了大量的蛋白质交互网络数据，它蕴含着蛋白质 之间相互作用的重要信息。利用蛋白质交互网络数据来预测具有显著生物功能模 块在生物功能基因组研究中具有重要的理论和现实意义。 在传统的生物功能模块检测中，往往把蛋白质交互网络当成一个普通的网络， 蛋白质对应网络中的顶点，交互对应其中的边。应用Jaccard 系数、边的中心性等 传统的网络度量来对蛋白质交互网络拓扑特性进行研究。而事实上蛋白质交互网 络是一个富含生物结构的网络，不能仅仅考虑其拓扑度量，更应该考虑其结构度 量。本文提出了一种基于结构的对称性度量DS(Degree-Symmetry) 。研究了对称性 度量DS 与蛋白质交互网络中的弱连接效应和拓扑相似性之间的关系。使用对称性 度量DS 建立了可信的虚拟网络，并在此基础上提出了一种“核－附件”生物功能 模块检测方法。 本文使用蛋白质交互网络数据进行了实验，实验结果表明对称性度量DS 相比 传统的网络度量能更好的刻画蛋白质交互网络中的弱连接效应。并且基于虚拟网 络检测出来的生物功能模块具有较高的准确性和较显著的生物意义。 关键字：蛋白质交互网络 络对称性 弱连接效应 模块检测 iv v Abstract In post-genomic era, the gene order has been decoded, but the biological function is still shrouded in mystery. Accompanied by the rapid development of biotic experiment, a large protein-protein interaction network database, which contains important information of the interaction between proteins, has been established. To predict the modules which have notable biological function by using protein-protein interaction network data is of great significance to the research of biological function genome. In traditional experiment of biological function modules, the protein-protein interaction network is always treated as a common network, in which the protein corresponds to the vertex and the interaction to the edge. Some traditional network measurement, such as The Jaccard coefficient and edge betweenness centrality, is applied in the research of topological characteristics in protein-protein interaction network. More than to consider only the topological measure of the protein-protein interaction network, the structural measure is more important to be taken into account, because the network contains abundant biological