Abstract
Emerging of microarray experiments have rendered simultaneous screenings of thousands of gene expressions possible, and subsequent analyses of transcription factors that highly correlate with differentially expressed genes can be conducted to determine the likely mechanism of gene regulation. Since transcription factors have been proven to work cooperatively with one another as composite elements by a growing body of evidence, previous efforts in identifying only one transcription factor for each experiment are clearly insufficient to decipher the complex underlying mechanism of expression changes. Therefore, the attempt here incorporates the analysis of TF binding positions and focuses on detecting TF pairs from genes with altered expressions. In combination of microarray and promoter binding site analysis, we predict pairs of transcription factors that mostly likely contribute to the regulated transcription of differentially expressed genes. The Expectation Maximization clustering of the TF binding positions is made available in order to detect a pattern in specific TF binding and predict the role it might play in regulation.