Based On Biochip Systematic Detecting Differencally Expressed Genes And Constructing Nonlinear Stability Network

Studying biological system with microarray data, it is crucial and basic to study biological system via detecting differentially expressed genes, selecting stable genes, working out the association between those genes. But there is no systematic approaches for studying biological system. We try to constuct a approach system to study biological system.Many methods, including parametric, nonparametric, and Bayesian methods, have been used for detecting differentially expressed genes based on the assumption that biological system are linear, which ignores the nonlinear characteristics of most biological systems. More importantly, those methods do not simultaneously consider means, variances, and high moments, resulting in relatively high false positive rate. To overcome the limitations, the SWang test is proposed to determine differentially expressed genes according to the equality of distributions between case and control. Our method latently incorporates functional relationships among genes to consider nonlinear biological system via utilizing information of skewness and kurtosis. To illustrate biological significance of high moments and how to indicate nonlinearity, we construct a nonlinear gene interaction model (NGIM), demonstrating that skewness and kurtosis could contain useful information of function association among genes in microarrays. It is important to ensure the initial expressed gene(s) in this model. Simulations and real microarray results show that false positive rate of SWang is lower than currently popular methods (T-test, F-test, SAM, and Fold-change) with much higher statistical power. Additionally, SWang can uniquely detect significant genes in real microarray data with imperceptible differential expression but higher variety in kurtosis and skewness. Those identified genes were confirmed with previous published literature or RT-PCR experiments performed in our lab.To deeply understand the NGIM, variable robust subset choosing (VRSC) have proposed to ensure initial changed gene(s) or variable(s). Results of statistical simulation display that the statistical power of VRSC is great, but its false positive rate is bad. To decrease the false positive rate of VRSC, we suggest that SWang test should firstly be used, then VRSC should be used.to select rubost varible subset.To deeper understand the nonlinear system or network, Kernel function Wang nonlinear correlation coefficient (KWNCC) have been developed based on NGIM, which incorporate the information of multiple moments simultaneously. Results of statistical simulation suggest that KWNCC could better measure the nonlinear correlation, comparing with the other six correlation methods.Analyzing three real gene microarray with those approaches and network analysis, the results suggest those approaches are effective for studying nonlinear gene interaction. Therefore, it is also built a analysis process for nonlinear systen or network. This will assist the development of bioinformatics, especially the analysis of microarray.