Research On Function Modular Mining Algorithm In Protein-Protein Interaction Network Based On Harmony Search Clustering Optimization

Protein-Protein Interaction(PPI)network is a network formed by interactions between proteins in organisms,it belongs to a kind of complex network and shows small world characteristics and scale-free properties.With the development of High-throughput biotechnology in recent years,more PPI data can be obtained.Mining function module in PPI is one of the new research topics in bioinformatics which has important guiding significance,it helps to predict unknown protein functions,and providing a theoretical basis for the study of diseases.Meanwhile,intelligent methods have applied widely in solving complex problems because of the superiority,the methods based on intelligent computing have been used in the PPI function module mining problem,which has become a new research hotspot.In this paper,the harmony search algorithm in the PPI function module mining problem is deeply studied,the main work includes:(1)An algorithm to detect function module in protein-protein interaction network based on harmony search clustering optimization(HMS-FMD)has been proposed,in order to improve the traditional harmony search algorithm search strategy,the method to find a set of nodes with larger aggregation coefficient in protein-protein interaction network is regarded as the objective function of the proposed algorithm.Through the analysis and comparison of the experimental parameters on the algorithm,the setting parameters of the optimal algorithm is presents,compared with other algorithms,the results show that this algorithm can effectively find out the function modules in PPI networks.(2)Current research generally the protein-protein interaction network as a side there is a deterministic undirected graph,but due to the high throughput biological detection technology of protein-protein interaction inherent error,the real of measured protein interaction is uncertain.On the uncertain graph,the computational complexity of protein function module mining problem is usually higher than the computational complexity of determining the same data mining problem.Base on the possible world model and the uncertainty of protein interaction networks,the algorithm of mining uncertain protein network function module harmony search optimization model was presents,through theoretical derivation,simplify computation complexity,using harmony search clustering optimization model,to search for the desired node density larger set as objective function.Through experimental analysis and comparison of the parameters of the algorithm,the algorithm has better clustering results.In this paper,the research on function modules detect in protein-protein interaction network based on harmony search clustering optimization enriches the theoretical research of PPI network function module detect algorithm,and provide theoretical guidance on the functional module mining of PPI network.