The Application Of Swarm Intelligence Algorithm In Mining Protein Complexes

In organisms,protein complexes to participate in the process of biological activity by the interactions between proteins,is the basis of ensure the normal order of the biological survival,and for the study of pathological process provides important biological theory basis.Therefore,the identification of protein complexes plays an important role in understanding cellular tissue and functional mechanisms.In recent years,with the rapid development and application of swarm intelligence algorithm in PPI network is increasing,not only in combination with the traditional method to identify the complexes,but also on the swarm intelligence algorithm itself.In this paper,the main research work includes two aspects,the Particle Swarm Optimization algorithm and the Glowworm Swarm Optimization algorithm is used in the clustering method called OPTICS algorithm based on density in design parameters and BFO algorithm clustering model to identify compounds,the specific work is as follows:(1)We introduce the basic principle and operation flow of Particle Swarm Optimization algorithm,the Glowworm Swarm Optimization algorithm and Bacterial Foraging Optimization algorithm,then describe the study of dynamic PPI network status and basic building,after that we apply the clustering method based on density to identify protein complexes in dynamic PPI network.(2)Due to the current PPI network data is imperfect and not accuracy,and protein complex have some special structures,the clustering algorithm based on density in the traditional way of distance calculation is not suitable for the calculation of the distance between the proteins in PPI network.So we using the calculation method of similarity between protein as a substitute for the calculation of distance in PPI network,the selection of calculation method can more accurately reflect the protein connected close degree between nodes.(3)Aimed at the core in the original OPTICS clustering algorithm,we redefine the core in PPI network,and use the swarm intelligence algorithm optimize OPTICS clustering algorithm,and by finding the optimal set of parameters to get the best clustering results.In the dynamic PPI network,swarm intelligence algorithm is used to optimize the clustering algorithm to improve the efficiency of the algorithm and find the global optimal value.The algorithm is tested in the four kinds of dynamic PPI data sets,and compared with the existing 7 kinds of methods and OPTICS algorithm,the experimental results show that the algorithm optimized by swarm intelligence algorithm is better than the comparison algorithm in the evaluation of f-measure and p-value.This means that the clustering results obtained from this algorithm have more biological significance than the protein complexes identified by other methods.(4)Aiming at the application of swarm intelligence algorithm in PPI network,we designed clustering model based on swarm intelligence algorithm.This article uses the bacterial foraging algorithm to design the clustering model,and use chemotaxis,reproduction,migration elimination and dispersal of bacterial foraging combined with core attachment structure of protein complex to identify the clustering module,then apply this model to the clustering on dynamic PPI network,make its reasonable and effective adjustment model,finally the better clustering results are obtained.We also tested on four data sets,and compare the algorithm with other 6 kinds of algorithms,the results show that the improved algorithm can effectively improve the recall ratio of the algorithm,and is superior to the contrast algorithm in the f-measure evaluation.