Summary Objective

The construction of genetic regulatory networks from time series gene expression data is an important research topic in bioinformatics as large amounts of quantitative gene expression data can be routinely generated nowadays. One of the main difficulties in building such genetic networks is that the data set has huge number of genes but small number of time points. In this paper, we propose a novel linear regression model for uncovering the relations among the genes. Methods The model is based on the multiple regression. It takes into account of the fact that the real biological networks have the scale-free property. Based on this property and the statistical tests, a filter can be constructed to filter some redundant interactions among the genes. By minimizing the distance between the observed data and the predicted data, the model can be finally constructed