Reconstruction of Gene Regulatory Networks using Differential Evolution

Gene Regulatory Network (GRN) is an abstract mapping of gene regulations in living cells that can help to predict the system behavior of living organisms. In this research, we use a model based inference method to reconstruct GRN from gene expression data. We use linear time variant model which is of particular interest among all other models because of its capability of discovering the non-linear interactions among genes in a reasonably short time even while dealing with noisy time-series data. Here, Differential Evolution (DE), a versatile, robust and well-known Evolutionary Algorithm (EA) has been used. The potency of the proposed method has been verified in gene network reconstruction experiments, varying the network dimension and characteristics, the amount of gene expression data used for inference, and the noise level present in gene expression profiles. Real expression dataset of SOS DNA repair system in Escherichia coli is used to reconstruct the regulatory network. All these experiments have proved the efficacy of the proposed reconstruction method.