Skip to Main Content
Recent technological progress on high-throughput measurements for gene expression such as microarray analysis enables us to collect time-series gene expression data for each of tens of thousands of genes. Although a genomic analysis with those data has identified key genes relating to various diseases, few results on estimation of gene regulatory networks with real microarray data are available so far. Recently, the immediately early response (1ER) genes upon epidermal growth factor stimulation in a human breast cancer cell line, MCF-7, have been identified in which time-course microarray data were measured during 90 minutes and 63 1ER genes were chosen from tens of thousands of genes by using statistical analysis. In this paper, we estimate the gene regulatory networks among the 63 1ER genes. To this end, we apply an estimation method based on a mixed logic dynamical modeling developed in an earlier study to the microarray data. However, the original method is executable for continuous gene expression time-series data whereas the real microarray time-course data have very few time points. In addition, some presetting parameters in the model are critical for a successful result on a network estimation. Then, we add a preprocessing and Monte Carlo-based calculation for die original method.