Integration of statistical inference methods and a novel control measure to improve sensitivity and specificity of data analysis in expression profiling studies

Abstract

Statistical methods have proven invaluable tools for enhancing the quality of microarray analysis. In this study, we used different methods such as significance analysis of microarrays (SAM) and Bayesian analysis of gene expression levels (BAGEL), to analyze the same set of raw data in an attempt to maximize the chance of identifying genes whose expression were significantly altered in gastric cancers. In addition, we examined the utility of an additional set of reference in controlling the variances and enhancing the quality of the results. Our results showed that BAGEL has the advantage of detecting small yet statistically significant differences, which might be of biological significance. Furthermore, introducing an additional control into the BAGEL, we were able to minimize the influence of the variances and significantly reduce number of potential false positive hits. BAGEL incorporates a novel control significantly improve the sensitivity and specificity of gene expression profiling analysis.