| Protein-protein interaction(PPI)network is a network of interactions between all proteins in an organism.The protein functional module is a collection of proteins in a PPI network that interact to perform a particular biological function.Mining protein functional modules in PPI networks is an important part of proteomics research.It not only helps people understand the operational mechanism of life at the molecular level,but also has great significance for disease diagnosis and development of new drugs.Therefore,in order to effectively detect the protein function modules in PPI networks,the computing methods based on clustering mechanism are rapidly developing.Among them,the solution method using the swarm intelligence optimization mechanism has shown good results.In this paper,some research work based on fireworks algorithm(FWA)are carried out from the following two aspects:(1)As a new swarm intelligence optimization algorithm,fireworks algorithm is possible to obtain better results in PPI network function module detection.Therefore,this paper proposes a method based on fireworks algorithm for functional module detection in PPI networks(FWA-FMD).Firstly,each firework individual is initialized as a candidate solution based on the label propagation idea by combining the topological and functional information.Secondly,in each generation of evolution,each firework individual is optimized by using explosion operation with local search and global search self-adjustment capabilities,and the next generation of fireworks individuals are selected by using elite retention and roulette strategy.Finally,the nodes with the same label in the optimal firework are divided into the same function module to obtain the final function module detection result.The experimental results on four public PPI network datasets of Saccharomyces cerevisiae and Homo sapiens show that the FWA-FMD algorithm not only costs less time than GA-PPI,ACC-FMD and BFO-FMD,but also has obvious advantages in many evaluation indicators compared with some classical algorithms,which can better identify functional modules.(2)To overcome the defect that FWA-FMD is easy to fall into local optimum,this paper proposes a method based on fireworks algorithm with differential evolution strategies for functional module detection in PPI networks(FWADE-FMD).This algorithm takes the fireworks algorithm as the basic framework,redesigns an explosion operation based on the influence of nodes,which makes full use of the neighborhood information of nodes in the PPI network.At the same time,three strategies of differential evolution algorithm are introduced: mutation,crossover and selection,which increases the diversity of population and enlarges the search scope.Firstly,this algorithm uses the population initialization method of FWA-FMD to initiate the population.Secondly,the explosion sparks are generated by the explosion operation based on the influence of the node,and the intermediate population is generated by the elite retention and roulette selection strategy.Then,the intermediate population generated by the fireworks algorithm is further optimized by the mutation operation,cross operation and selection strategy of the differential evolution algorithm.Finally,the function module detection result is obtained by decoding the optimal individual in the population.The experimental results on four PPI network datasets show that the FWADE-FMD algorithm can not only obtain better solution performance than the FWA-FMD algorithm,but also has obvious competitiveness in overall performance compared with some classical algorithms. |
