Fractional order transcriptional regulation model for Hela cell BPI gene of cervical cancer
Ganggui Zhang, Ruirui Ji, Meng M. Zhao · 2021 China Automation Congress (CAC) · 2021
The treatment and research of cervical cancer are the hot issue in medical care. A reasonable mathematical model is helpful for revealing the regulatory mechanism of cancer genes. This paper studied a new method to reconstruct transcriptional regulation model with fractional differential equation. Both model error and observation error are considered to improve the model accuracy, then multi-objective particle swarm algorithm is applied to identify the model order and parameters together, in which Adam-Bashforth-Moulton method is used to solve the fractional model. The results on simulation experiments prove that the proposed multi-objective optimization algorithm has higher identification accuracy than single-objective optimization. The experimental results on the BPI gene data of HeLa cells in cervical cancer show that fractional order model proforms more reliably for real data than integer order model, and the multi-objective optimization algorithm is more robust for model inference.