Learning the Structure of Bayesian Networks Representing Influence Relations among Genes

A Bayesian network is a graph-based model of joint multivariate probability distributions that captures properties of conditional independence between variables. Such models are an effective way to characterize probabilistic and causal relations among variables providing a clear methodology for learning from observations. In recent years their use to recover transcriptional regulatory networks from static microarray data is becoming an active area of bioinformatics research. The intent of this paper is to provide a review on structural learning of Bayesian Networks and to compare described methods on a benchmark dataset, the Hepatic Glucose Homeostasis network, that describe results of microarray experiments.