Advances in microarray technology have resulted in an exponential rise in gene expression data. Partially as a result of this, full genome sequences have been reported for many organisms. In addition several methods have been developed (a) for assigning functionality to previously unknown genes and (b) for measuring the output (i.e. gene expression) of the gene regulatory network. The knowledge of the gene regulatory network further gives insights about gene pathways. This information leads to many potential applications in medicine and molecular biology, examples of which are identification of metabolic pathways, complex genetic diseases, drug discovery and toxicology analysis. Also, gene regulatory networks allow comparison of expression patterns of many uncharacterized genes; this comparison provides clues to gene function. A variety of models (such as neural networks, Boolean networks, and Bayesian) have been proposed in recent times. Although each of these models have individual strengths, none of them addresses important issues such as time delay, or make use of available biological information. In the work presented here we demonstrate that networks can efficiently model natural biological processes, specifically gene regulatory systems. Through the modeling approach, we have inferred gene regulatory networks using a time course data set for (a) lambda bacteriophage infection, (b) osteoblast study, and (c) rat central nervous system (CNS) development. The results compare favorably with experimental results from the literature. (c) 2006 Elsevier Ltd. All rights reserved.