Introducing Parameter Clustering to the OED Procedure for Model Calibration of a Synthetic Inducible Promoter in S. cerevisiae

In recent years, synthetic gene circuits for adding new cell features have become one of the most powerful tools in biological and pharmaceutical research development. However, because inherent non-linearity noisy experimental data, experiment-based model calibration these parts is perceived as a laborious time-consuming procedure. Although optimal design (OED) based on Fisher information matrix (FIM) has been proved to be an effective means improve efficiency, required calculation increases dramatically with size (parameter number). To reduce OED complexity without losing accuracy, this paper proposes two approaches different parameter clustering methods validates accuracy calibrated models in-silico experiments. A inducible promoter S. cerevisiae adopted bench-marking. The comparison traditional off-line approach suggests that both significantly problems (for at least 49.0%), while slightly improving (11.8% 19.6% lower estimation error average FIM-based sensitivity-based approaches). This study implicates calibrating non-linear pathways, cluster-based could beneficial efficiency design.