Extraction of interaction information among genes from gene expression time series data

Gene regulatory network gives the idea about the nature of interaction among the genes present in the DNA of a living species. Detection of gene regulatory network from gene expression data is of prime interest to the researchers. This paper considers modelling of the gene regulatory network identification problem using a fuzzy recurrent neural network, and obtains the interaction weights among the neuron using differential evolution algorithm. A cost function is designed, the minimization of which yields the solution to the problem. In order to improve the solution further, a heuristic based local search is proposed. Computer simulation of the proposed inference algorithm revels that it is able to predict the signs of all the existing weights accurately.