A Mixture Model Approach to Detecting Di erentially Expressed Genes with Microarray Data

An exciting biological advancement in the last few years is the use of microarray technologies to measure simultaneously the expression levels of thousands of genes The bottleneck now is how to extract useful information from the resulting large amounts of data An important and common task in analyzing microarray data is to identify genes with altered expression under two experimental conditions We propose a nonparametric statistical approach to handle the problem when there are a small number of replicates under each experimental condition Speci cally we propose to estimate the distributions of two t statistic type scores related with gene expression levels using Normal mixture models A comparison of these two distributions by means of a likelihood ratio statistic is used to identify genes with signi cantly changed expression where the cut o point for statistical signi cance is determined using a parametric bootstrap technique which also controls the number of genes incorrectly identi ed The methodology is applied to a data set containing expression levels of genes of rats with and without pneumococcal middle ear infection