^(13)C代谢通量估计的量子粒子群优化算法  

Quantum-behaved particle swarm optimization algorithm for ^(13)C metabolic flux analysis

作　　者：吕丽丽[1] 孙俊[1] 须文波[1] 余永红[1] 

机构地区：[1]江南大学信息工程学院,江苏无锡214122

出　　处：《计算机与应用化学》2011年第2期151-155,共5页Computers and Applied Chemistry

基　　金：国家自然科学基金项目(60474030)

摘　　要：^(13)C标记实验的代谢通量分析(^(13)CMFA)是探索代谢网络的重要途径。^(13)C通量估计是以碳富集度平衡为条件的全局优化问题,带有众多约束条件和存在多个局部极小点等特点,如何高效地求解是^(13)C MFA中的难点,也是实现通量精确估计的关键。量子粒子群优化算法显著特点是控制参数少,设置简单,具有较好的全局搜索能力,适应于通量估计。本文提出量子粒子群优化算法结合最小二乘计算求解噪音环境下的环磷酸戊糖代谢网络的通量,以带约束的最小化问题为目标优化函数,仿真实验验证了量子粒子群优化算法是1种有效的通量估计分析算法。Metabolic flux analysis with measurement data from ^（13）C tracer experiments has been an important approach for exploring metabolic networks.In this paper,flux estimation is formulated as a global optimization problem by carbon enrichment balances.Characteristics such as existence of multiple local minima and non-linear optimization make this problem a special difficulty,quant-um-behaved particle swarm optimization algorithm,which is a robust and convergence fast optimization algorithm,is applied to flux estimation.The algorithm performances are illustrated and compared with ordinary least squares estimation through simulate on the cyclic pentose phosphate metabolic network in a noisy environment.It is shown that quantum-behaved particle swarm optimization algorithm,is an efficient approach for flux quantification.

关 键 词：量子粒子群算法 代谢分析 最小二乘估计 (13)C标记实验 

分 类 号：N945[自然科学总论—系统科学]