Abstract

The potential of microarray experiments to profile the expression of thousand genes simultaneously has focused the interest of scientists during last twenty years. Microarray experiments provide valuable information about how genes compete and are associated to produce complex responses and cooperative effects. The modeling and analysis of such rich data sets has attracted researchers from different fields such as Statistics, Data Mining or Signal Processing. The results of this kind of analysis is a good candidate to undercover valuable information regarding the diagnosis, prognosis and drug design for specific diseases. In this work, we have revised recent patents that address the problem of model and infer the gene regulatory network from a microarray time-series data sets. Moreover we have compared these approaches with an inference method developed by the authors. Our methodology establishes a novel approach that combines a Markov linear model with the variational-Bayesian framework to undercover the Gene Regulatory Network.

Keywords: Boolean network, bayesian network, gene regulatory networks, microarrays, DNA macromolecules, GRN, cell-cycle network, cycle-regulated genes, boolean networks, Miyano's differential model, Variational Bayesian Maximization, Variational Bayesian Expectation, Kullbar-Laibler divergence, Bayesian formulation, GRN topology