Parameter Estimation of Gene Regulatory Network Using Honey Bee Mating Optimization

Abstract

Computational biology is the esteemed interdisciplinary field where expertise from the fields like Mathematics, Statistics and Computer Science are applied to have the insight in biological phenomenon. Advanced methods and techniques of biotechnology and allied fields facilitate the availability of biological data to the researchers of computational biology. Microarray time series gene expression data is such an effective dataset which uncovers the regulatory relationships between any pair of genes in a gene set and hence facilitates the reconstruction of Gene Regulatory Network. The Artificial Neural Network environment is used to find the expression level of a gene at time t+?t in terms of the available expression level at time t. The underlying network parameters are uncovered as the simulated time series are compared with available real dataset in successive iterations. Estimation of the parameters of gene regulatory network is an important research area to be addressed. Here in this paper, the parameters are estimated using Honey Bee Mating Optimization algorithm. The intelligence of queen bees of the bee colony to select prospective drones for mating, crossover and mutation to support effective new genotypes and nurture of the good broods by worker bees is applied to solve the optimization problem of Parameter Estimation. Two experiments are conducted here. In experiment 1, the simulation based on the synthetic dataset of predefined parameters showed good performance accuracy. In the case of experiment 2, where real dataset was used, the cost convergence indicates the excellence of Honey Bee Mating Optimization in Parameter Estimation of Gene Regulatory Network.