Clustering Analysis Of Gene Expression Profile Data Based On Meta-heuristic Algorithms

In this era of explosive growth of biological information,the technology of biclustering analysis of gene expression profile data has gradually become a research hotspot.At the same time,various biclustering algorithms are being continuously developed to determine the co-expressed genes in gene expression data.In order to enrich the biclustering algorithms and improve their applicability,this paper proposes a new biclustering algorithm-Binary Artificial Fish Swarm Algorithm(BAFSA),which is an improved meta-heuristic search algorithm,and combines traditional Artificial Fish Swarm Algorithm(AFSA)with the binary form.Because this algorithm uses a fitness function based on linear correlation,it can find genes with shift and scale patterns,which makes up the shortcomings of traditional clustering algorithms.In this paper,the algorithm is applied to the actual genetic datasets,which can extract biologically significant biclustering subsets with the good clustering performance.Nowadays,the traditional biclustering algorithms have two drawbacks:(1)it may easily fall into a local optimum;(2)it is only applicable to some specific datasets.In order to make the BAFSA algorithm proposed in this paper be able to jump out of local optimum and have wide applicability,this paper makes innovative improvements to this algorithm.This paper combines the binary artificial fish swarm algorithm(BAFS)with the binary simulated annealing algorithm(BSA),and proposes a new hybrid algorithm BAFS-BSA-BIC.When this new biclustering method is applied to multiple data sets,many biologically significant biclustering subsets are searched,showing the good clustering performance.In addition,compared with other classic biclustering algorithms as well as the original BAFSA algorithm,it can be found that the robustness of the algorithm and the quality of the search biclusters are better than those of the algorithms.