doiSerbia

A novel network aligner for the analysis of multiple protein-protein interaction networks

Chen Jing (School of Artificial Intelligence and Computer Science, Jiangnan University Wuxi, China + Jiangsu Provincial Engineering Laboratory of Pattern Recognition and Computing Intelligence, Jiangnan University, Wuxi, China), chenjing@jiangnan.edu.cn
Huang Jia (School of Artificial Intelligence and Computer Science, Jiangnan University Wuxi, China)

The analysis of protein-protein interaction networks can transfer the knowledge of well-studied biological functions to functions that are not yet adequately investigated by constructing networks and extracting similar network structures in different species. Multiple network alignment can be used to find similar regions among multiple networks. In this paper, we introduce Accurate Combined Clustering Multiple Network Alignment (ACCMNA), which is a new and accurate multiple network alignment algorithm. It uses both topology and sequence similarity information. First, the importance of all the nodes is calculated according to the network structures. Second, the seed-and-extend framework is used to conduct an iterative search. In each iteration, a clustering method is combined to generate the alignment. Extensive experimental results show that ACCMNA outperformed the state-of-the-art algorithms in producing functionally consistent and topological conservation alignments within an acceptable running time.

Keywords: graph data analysis, big data, protein-protein interaction network, network clustering, seed-and-extend strategy