Krishnadev O, Brinda K V, Vishveshwara Saraswathi
Molecular Biophysics Unit, Indian Institute of Science, Bangalore, India.
Proteins. 2005 Oct 1;61(1):152-63. doi: 10.1002/prot.20532.
We present a simple method for the analysis of large networks based on their graph spectral properties. One of the advantages of this method is that it uses a single numerical computation to identify subclusters in a connected graph, which can significantly simplify the complexity involved in analyzing large graphs. This is illustrated using a network of protein chains constructed on the basis of their structural similarities. The large-scale network properties and the cluster and subcluster organization of the protein chain network are presented. We summarize the results of structural and functional analyses of the nodes present in these clusters and elucidate the implications of structural similarity in the protein chain universe.
我们提出了一种基于图谱特性分析大型网络的简单方法。该方法的优点之一是，它使用单一数值计算来识别连通图中的子簇，这可以显著简化分析大型图所涉及的复杂性。这一点通过基于蛋白质链结构相似性构建的网络得到了说明。本文展示了蛋白质链网络的大规模网络特性以及簇和子簇组织。我们总结了这些簇中节点的结构和功能分析结果，并阐明了蛋白质链世界中结构相似性的意义。
