A new method for perturbation experimental design in gene regulatory network identification

Inferring gene regulatory networks by high-throughput data is a fundenmental problem in systems biology. The interactions between genes, proteins and other small molecules are typically described by gene regulatory networks, which are nonlinear and sparce. We linearize the nonlinear system of the segmentation polarity network of Drosophila melanogaster and infer the interaction between genes in the network by perturbation experimental data. The genes expression level are measured by microarray experiments. we calculate the parameters' changes forced by inputs of the experiment, and give a new method for experimental design in which the inputs facilitate precise estimation of the parameters. All the data in calculation is simulated in silico.