Research On Protein-protein Network Mining Based On Clustering Coefficient Overlapping Community Detection

With the perfection of human genome sequencing,proteomics has become an important research content in life sciences.Due to the rapid progress of high-throughput technologies,the data of protein interactions are increasing constantly,and researchers are no longer limited to the collection and management of data,but to pay attention to the meaning and value behind these biological data.The study of protein-protein interaction networks(PPIN)help to predict the functions of unknown proteins at the molecular level,thereby further revealing the regularity of cellular activities.The research of functional modules in PPIN is an important way to further understand the mechanism of action and organization principles of life activities.In order to detect functional modules and analyze overlapping proteins between modules,we propose the NLC algorithm based on the topological structure of PPIN,which overcomes the low recognition rate of overlapping modules and improves the accuracy of the division results.The research includes the following:(1)The algorithm in this paper considering the influence of neighbor nodes when calculating the quality of nodes,and introduces the local clustering coefficient of neighbor nodes to measure the quality of nodes to obtain more reasonable seed nodes.(2)When measuring the distance between edges,the algorithm combines the shortest topological distance and Jaccard distance between edges to improve the measurement method of the distance between edges,which can be used to measure the distance between non-adjacent edges.(3)Aiming at the phenomenon of excessively overlapping nodes in the communities,the NLC algorithm formulates a community optimization strategy,which corrects the excessively overlapping nodes in the original division,so that the result is closer to the real division.(4)Through experiments on three types of real networks,LFR artificial networks and protein-protein interaction networks,and compared with LC,CPM,CES,CNS,the results show that the NLC algorithm is more accurate.It is verified that the NLC algorithm can not only find reasonable community structures but also find overlapping community structures in complex networks.Finally,functional enrichment analysis was used to verify whether the modules found in PPIN had biological significance,and it was concluded that the protein functional modules detected by the algorithm in this paper had biological significance.The above is the content of this paper.The experimental results show that the algorithm NLC in this paper has achieved good performance and improved in various evaluation indicators.It verifies that the NLC algorithm can be used for module detection in social networks and functional module detection in biological networks.