Determining and analyzing differentially expressed genes from cDNA microarray experiments with complementary designs

We present a hierarchical Bayes model for determining genes in the sporulation pathway of Bacillus subtilis (B. subtilis) using two complementary replicated cDNA microarray experimental designs. The first design involves the mutation of a transcription factor, Sigma factor E (σ^E), and the second is an overexpression of this factor. We first normalize the microarray data using a rank invariant method. Genes found to be overexpressed in both experimental designs are further examined experimentally to determine their role in the sporulation pathway. Through statistical and experimental methods we found 181 genes that had not been previously described as controlled by σ^E. We identify the chromosome locations of clusters of σ^E-controlled genes using a nearest neighbor scan-statistic, and determine B. subtilis functional categories that are over-represented in subsets of expressed genes.