Analysis of Module of Co-Regulated Genes by Integrating Information from Sequence and Microarray

The researches of module of co-regulated genes act as important role in research of gene regulatory networks. Present analysis of module of co-regulated genes generally have 2 problems: The first one is that the co-expressed genes are generally clustered as the co-regulated genes, but recent studies show that only 50% genes are co-regulated when their Spearman correlative coefficients are bigger than 0.84, so the co-regulated genes need to be refined from cluster of co-expressed genes; the other problem is that the common transcription factor (TF) is usually unknown, this makes obstacle in quantitative analysis on regulation relationships between TF and target genes. This paper presents a method for analysis of module about co-regulated genes by integrating information from sequences and microarray data. Firstly we cluster the genes by the similarity of expression profiles; then the co-regulated genes are refined by analysis of transcription factor binding sites (TFBS) in their promoter regions; the expression level of the gene corresponding to the common transcription factor can be estimated by Viterbi decoding based on Hidden Markov Model (HMM), whose parameters are learned by Baum-Welch algorithm. By consulting the related articles of experiments and standard database, the correctness and effectiveness of our method are validated.