Optimize observational time points to maximize the inferability of gene networks

In the post-genomic era methods that are capable to estimate gene networks, e.g., signalling, metabolic or transcriptional regulatory networks, from high-throughput data are of utmost interest. In this paper we study the inferability of gene networks from time series data of a simulated gene network mimicking microarray data. The stochastic model we apply is capable of reproducing characteristic features of experimental data and, hence, can serve as a biologically plausible model of gene expression. The major purpose of this paper is to investigate the influence of the selected observational time points on the inference of a gene network. More precisely, we aim to find the optimal observational time points P of a long time series T >> P that lead to a maximization of a scoring function evaluating the quality of the estimated gene network structure.