Model tree to improve the inference of gene association networks

Abstract

We propose a novel methodology to infer gene association networks from gene expression profiles (microarray data) based on the application of model tree. We first build a regression tree for each gene and second, we build a graph from all the linear relationships among output and input genes taking into account whether the pair of genes is statistically significant. Then, we apply a statistical procedure to control the false discovery rate. Part of this methodology is a key component in a prediction- based method for a cardiovascular problem based on the discovery of clinically relevant transcriptional association networks. The aim of this second method is to apply the information encoded in gene networks for prognostic purpose which is one of the crucial objectives of systems biomedicine.