Skip to Main Content
The correct inference of genetic regulatory networks plays a critical role in understanding biological regulation in phenotypic determination and it can affect advanced genome-based therapeutics. In this study, we propose a joint evolutionary algorithm and H∞ filtering approach to infer genetic regulatory networks using noisy time series data from microarray measurements. Specifically, an iterative algorithm is proposed where genetic programming is applied to identify the structure of the model and H∞ filtering is used to estimate the parameters in each iteration. The proposed method can obtain accurate dynamic nonlinear ordinary differential equation (ODE) model of genetic regulatory networks even when the noise statistics is unknown. Both synthetic data and experimental data from microarray measurements are used to demonstrate the effectiveness of the proposed method. With the increasing availability of time series microarray data, the algorithm developed in this paper could be applied to construct models to characterize cancer evolution and serve as the basis for developing new regulatory therapies.