The Constrction Of Online Analysis System Based On Gene Expression Data

The transcriptome is, at specific development stages or physiological conditions, a set of all transcrption segments in a cell. Transcriptome study is a very important mean for clarifying the function areas of genome, revealing the molecular compositions of cells and tissues, and understanding the occurrence and development of diseases. In recent years, with the development of microarray and high-throughput sequencing technology, transcriptome research produces massive expression data. Understanding their biological significance has become a hot research field of bioinformatics and computational biology.Through integrating common analysis tools and optimizating high efficient data mining algorithm of gene expression profile, we developed an online expression profile analysis system named FOGA(Factory Of Gene Analysis) based on technologies related to Java paltform. At present, the basic functions of the FOGA have included: the analysis of the differentially expressed genes, the clusterring of genes and samples, the visualization of genes located on chromosomes, the ordering of genes based on importance, the constructing of gene regulatory networks etc.. FOGA is highly flexible for data inputs. It can accept gene expression data produced by the platforms of high-throughput, microarray etc.. It completes the anlysis tasks according to parameters user defined and outputs the visualization results of analysis friendly. At the same time, FOGA has a friendly user interface, which makes it more efficient and convenient than the way of traditional scripts analysis. It uses the mode of modular design, which lays a good foundation for future expansion of new features. Moreover, FOGA implements a function of gene ranking based on the theory of Joint ι2,1-Norms Minimization, which can screen the genes related to diseases. Using the expression data of a set of samples of DBA patients with bone marrow for testing FOGA system, we found a large number of pathogenic genes associated with DBA. Compared with the traditional method of obtaining the key genes by expression, interaction networks and other complex means, FOGA is more intelligent and efficient. It opens up a new way for the study of functional genomics.