Reverse engineering gene regulatory networks

Abstract Complex networks of genes, proteins, and small molecules interact to determine cellular function. The reverse engineering or inference of gene regulatory networks using DNA sequence data, protein–DNA binding data, and observed molecular abundance data has become a major interest of the biological community. We discuss recent successes and remaining challenges in the construction of gene network models and their use for gaining biological insight. We draw lessons from the detailed discussion of several recent gene network inference studies, focusing on the novel advances and limitations of each approach. These approaches differ in the strategies used to incorporate biological data, the level of physical detail and type of model used, and the purpose of the modeling in terms of biological discovery. The approaches we discuss aim to gain insights into the logic of combinatorial regulation occurring at a gene promoter, the activity of regulatory factors distinct from their abundance, the identification of key regulators in a gene network, and the prediction of drug compound mode of action. We suggest possibilities for future directions in network inference studies.