非线性规划在生物代谢仿真过程中的应用  

Application of Nonlinear Programming for Simulation of Biological Metabolic Process

作　　者：潘多涛[1] 史洪岩[1] 黄明忠[1] 袁德成[1] 

机构地区：[1]沈阳化工大学辽宁省化工过程控制技术重点实验室,沈阳110142

出　　处：《控制工程》2014年第6期896-899,903,共5页Control Engineering of China

基　　金：国家863计划项目(2008AA042902);国家自然科学基金资助项目(60874057)

摘　　要：随着系统生物学研究的不断深入,生物过程网络的规模逐步拓展,其模型的复杂程度也越来越高,数值积分等常规方法很难求解生物代谢网络这类大规模复杂动态系统,客观上需要新的分析方法去应对。针对典型的酿酒酵母糖酵解途径的动态模型,首先利用Matlab进行常规的数值积分求解,以此作为参照,提出将糖酵解的动态模型离散化后利用非线性规划解题器IPOPT进行动态仿真的思路,采用不同的离散步长,计算得到不同的仿真结果。计算结果显示,通过调整计算参数,IPOPT可得到与参照一致的仿真结果,这表明非线性规划可作为分析优化生物过程这类大规模复杂动态系统的新方法,为下一步进行模型参数估计、多目标动态优化等打下基础。With the development of research on systems biology, the biological process network scale gradually expanded and the model in- creasingly became more complex. It was difficult to solve the biological metabolic network of large - scale dynamic complex systems by conven- tional methods. Objectively, it required a new analysis method to deal with. Aimed at the dynamic model of glycolysis pathway of Saccharomy- ces cerevisiae, the model was solved by MATLAB via conventional numerical integration method, and the results were taken for control subject. After discretization the dynamic model of glycolysis was solved by IPOPT which was a famous nonlinear programming solver. The simulation results were compared by using variable step of discretization. The results were shown to be consistent with the control subject, and nonlinear programming could be used as a new approach for the optimization of biological process and parameter estimation.

关 键 词：系统生物学 代谢途径 糖酵解 计算机仿真 非线性规划 

分 类 号：TP311.52[自动化与计算机技术—计算机软件与理论]