Research On Recognition Method Of Complex In Protein Networks

In recent years,along with the development of proteomics and systems biology research gradually thorough,the protein interaction network recognition complex has gradually become a hot research topic in bioinformatics.High throughput data to produce a large number of experimental data is bound to the protein,contains a large number of false positives,false positive data will affect the final effect calculation method of complex search.Therefore,the calculation method of complex search prior to the need to remove the false positive effect.This paper makes a study on the interaction of protein networks in protein,are calculated by using the clustering method to search for protein complexes in which the main points of research include the following two points:(1)In order to reduce influence of the false positive discovered through high-throughput data experiment of protein data,so as to enhance the search for protein complex identification accuracy,this paper proposes a novel search protein complex algorithm:MCLAd.First,based on the characteristics of the protein interaction network topology,choose to deal with network topology method AdjstCD for the high-throughput protein mutual data pretreatment,data reduction of false positive;secondly,combined with classical MCL clustering algorithm clustering identification of protein complexes;finally,will our algorithm MCLAd and existing seven excellent methods:MCL,ClusterONE,MCODE RRW,RNSC,coach and core algorithm respectively are compared with the experiments.Through a large number of experiments show that the proposed search algorithm MCLAd experimental effect is very obvious,there are very good experimental performance to search protein complexes.(2)Advantages of three methods of MCL and RRW integrated ClusterONE,we put forward the has been new protein complex search the MCR algorithm,and the MCR respectively with MCL,coach,ClusterONE,MCODE RRW,RNSC,core and our proposed MCLAd method for experimental comparison.We will be the algorithm respectively in biogrid protein mutual role of database and dip protein mutual database,the two data sets,and will search for income of the protein complexes respectively with SGD and the MIPs both complex standard data sets for experiments.Experimental results show that compared with other search protein complex algorithm,MCR algorithm to search for a protein complex with the standard complex data set matching,above the search efficiency and reliability has very big advantage.