Mining Conserved Topological Structures from Large Protein-Protein Interaction Networks

Analysis of Protein-Protein Interaction (PPI) networks is of great significance in evolutionary biology. Because of high computation cost, recently multi-PPI network alignment becomes hot topic. In this paper, we proposed conserved topological structures mining based multiPPI network alignment technology. The most challenging problems in conserved topological structure mining are the large size of the real PPI networks and the requirement of inexact alignment of networks. To solve this problem, we develop an algorithm framework: Conserved Substructure Mining (CSMiner) for PPI networks. In the algorithm, we synthesize many techniques to boost the mining procedure, including a novel efficient pattern growth method, an efficient node disjoint subgraph homeomorphism determination algorithm and the integrated strategy of these two algorithms. We also demonstrate the efficiency and effectivity through the experiments on real PPI networks of Saccharomyces Cerevisiae and Drosophila Melanogaster. Supported by the National Natural Science Foundation of China under Grant No.60303008 and 60373133; the National Grand Fundamental Research 973 Program of China under Grant No. 2005CB321905