Research On The Identification Of Key Nodes In The Complex Network

With the in-depth study of complex networks, key nodes research which plays an important role in complex network research is gaining more and more attention. Study of key nodes in complex networks has a lot of practical significance, for example, by analysis of the key nodes in the Internet can help improving network performance, increase network survivability; through customer relationship network node analysis can find key customers;key nodes analysis in protein interaction network can identify key disease genes. Besides, it has broad application in drug development, sales, and other aspects of the detection of cases.A lot of networks have been proved to be complex networks, such as social networks,communication networks and protein interaction networks. Some nodes throughout the network play key roles in these networks. Study of key nodes in complex networks is important for identifying criminals, the key disease genes, and key influential users in social network. How to design an efficient algorithm to identify key nodes of complex networks is currently a hot research area of computer application. In this paper, taking the identification of key nodes in protein interaction network as an example, we examined methods of recognition key nodes in complex networks.Currently, numbers of key protein identification methods have been proposed, such as methods based on node degree. In this kind of research, the size of node degree of as the stand for measuring node importance. Other methods based on the vertex betweenness, in such processes the vertex betweenness or edge betweenness as a measure of the importance of protein. Some methods are based on the node importance evaluation algorithm similarity contributions, as well as the importance of the node field theory model based evaluation algorithm. Page Rank algorithm is an important algorithm using for page ranking which has been modified for node importance measure in protein-protein interaction network. Based on the protein directly and indirectly to measure the strength of the interaction between proteins and reconstruct network evaluation algorithm was proposed which was done based on the importance of protein nodes. Comparion analysis of the key nodes identified by our methods with that of other methods, results show that the proposed method has a higher accuracy.Identification of key proteins node is important. On the one hand we can find functionally similar protein; it can also assist predicting unknown protein function, in particular, has an important role in disease diagnosis and treatment, for the development of new drugs. Besides it has important value in drug target discovery. In addition, it also has important significance in understanding the pathogenesis of complex diseases.