A new hybrid bee evolution algorithm for parameter estimation in biological model

The development of accurate and reliable models for biological systems plays an important role in both systems and synthetic biology. The models are constructed based on the ordinary differential equations to observe the concentration change of specific biochemical products. These formulations usually depend on a set of parameters that reflect the physical properties of the systems, such as reaction and kinetic rates. In most cases, these parameters are estimated by fitting the model prediction with the corresponding experimental data. Due to the noisy and incomplete experimental data, metaheuristics methods are utilized to find the optimal values of these parameters by minimizing the difference between both data. In this paper, a new optimization method is introduced for the biological model parameter estimation. The proposed method is developed based on the hybridization of Artificial Bee Colony (ABC) and Differential Evolution (DE) methods. In general, this method employs the evolutionary operations of the DE method to improve the neighboring searching strategy via the ABC method. The accuracy and reliability in estimating the parameters are demonstrated by using a model of lactose feedback regulation in a bacterial cell. The results showed that the performance of the proposed method has outperformed the existing methods.