UNCOVERING GENE REGULATORY NETWORKS USING VARIATIONAL BAYES VARIABLE SELECTION

摘要

In this paper, we propose a Bayesian approach for reconstructing gene regulatory networks (GRNs) based on microarray data. We focus on a variable selection formulation and develop a solution by a variational Bayes Expectation Maximization (VBEM) learning rule. The major advantage of the VBEM solution over Monte Carlo sampling based approach is its lower computational complexity. This makes it appealing for uncovering large networks.