Development of a system for the inference of large scale genetic networks.

We propose a system named AIGNET (Algorithms for Inference of Genetic Networks), and introduce two top-down approaches for the inference of interrelated mechanism among genes in genetic network that is based on the steady state and temporal analyses of gene expression patterns against some kinds of gene perturbations such as disruption or overexpression. The former analysis is performed by a static Boolean network model based on multi-level digraph, and the latter one is by S-system model. By integrating these two analyses, we show our strategy is flexible and rich in structure to treat gene expression patterns; we applied our strategy to the inference of a genetic network that is composed of 30 genes as a case study. Given the gene expression time-course data set under the conditions of wild-type and the deletion of one gene, our system enabled us to reconstruct the same network architecture as original one.