The Research And Implementation On Clustering Algorithm Of Gene Expression Data

With the rapid development of gene-chip technology and advanced biotechnology gene-chip can measure and analysis a large number of gene expression profiling rapidly at the same time, which further accelerated the emergence of gene expression data. How to effectively analysis,organize and deal with these vast amounts of gene expression data extracted from an effective biological, medical information has become a hot spot of concern and research. Clustering as one of the main technologies of gene expression data analysis has a wide range of applications in study of common functions, interaction and synergy control of gene. At present, there are a lot of clustering algorithm used in the clustering of gene expression data.This paper apply quantum-behaved particle swarm optimization algorithm (QPSO) to the gene cluster analysis, which constitutes QPSO gene clustering algorithm. The paper shows the superiorities of QPSO gene cluster algorithm by compared with other cluster algorithms through experiments .The main work of this paper as follows:(1) Based on the QPSO algorithm, we use another kind of fitness function TWCV instead of its original commonly used objective function and constitute QPSO gene clustering algorithm. The QPSO algorithms with a new fitness function applied to clustering on gene expression datas can avoid situation that gene vector classificated in non-equilibrium.(2) Combining the advantages of K-means and QPSO,PSO clustering algorithm, it proposed the other new clustering algorithm KQPSO and KPSO. And compare them with genetic clustering algorithm QPSO in gene expression data on the advantages and disadvantages of clustering through experiments(3) Using the optimize the overall characteristics of genetic algorithms (GA) and the fast convergence properties of K-means formats a fast genetic algorithm (FGKA).We apply it in the gene expression clustering and to comparie the different characteristics of FGKA and QPSO gene clustering algorithm in gene expression data clustering through experiments.(4) Applying DGQPSO algorithm to clustering gene expression data and comparing the performances to QPSO algorithm on gene expression data.At the last of this paper, it makes a system interface and gives a brief introduction about the functions,system interface of each module in the software system of GCS which will developed in future.