Novel metaheuristic for parameter estimation in nonlinear dynamic biological systems.
Novel metaheuristic for parameter estimation in nonlinear dynamic biological systems.
复制标题
非线性动力学生物系统参数估计的新型元疗法。
DOI：
10.1186/1471-2105-7-483
复制
发表时间：
2006-11-02
影响因子：
3
通讯作者：
Banga, Julio R
中科院分区：
文献类型：
--
作者：
Rodriguez-Fernandez, Maria;Egea, Jose A;Banga, Julio R
我们考虑生物系统非线性动态模型中的参数估计（模型校准）问题。由于许多此类问题频繁出现病态和多模态，传统的局部方法通常会失败（除非通过对参数向量的很好的猜测进行初始化）。为了克服这些困难，全局优化（GO）方法被建议作为可靠的替代方案。目前，确定性 GO 方法无法在合理的计算时间内解决此类实际大小的问题。相比之下，某些类型的随机 GO 方法已经显示出有希望的结果，尽管计算成本仍然很大。 Rodriguez-Fernandez 及其同事提出了混合随机确定性 GO 方法，该方法可以将计算时间减少一个数量级，同时保证鲁棒性。我们的目标是在不失去鲁棒性的情况下进一步减少计算量。我们开发了一种基于分散搜索方法的新程序，用于任意（甚至未知）结构的动态模型（即黑盒模型）的非线性优化。在这篇文章中，我们受运筹学领域最新发展的启发，描述并应用了这种新颖的元启发法来解决一系列复杂的识别问题，并与之前（上述）成功的方法进行了批判性比较。稳健而有效的参数估计方法在系统生物学及相关领域至关重要。本文提出的新元启发法旨在通过采用全局优化方法来确保这些问题的正确解决，同时将计算量保持在合理的值以下。这种新的元启发法应用于非线性动态生物系统的一组三个具有挑战性的参数估计问题，其性能显着优于以前用于这些基准问题的所有方法。
We consider the problem of parameter estimation (model calibration) in nonlinear dynamic models of biological systems. Due to the frequent ill-conditioning and multi-modality of many of these problems, traditional local methods usually fail (unless initialized with very good guesses of the parameter vector). In order to surmount these difficulties, global optimization (GO) methods have been suggested as robust alternatives. Currently, deterministic GO methods can not solve problems of realistic size within this class in reasonable computation times. In contrast, certain types of stochastic GO methods have shown promising results, although the computational cost remains large. Rodriguez-Fernandez and coworkers have presented hybrid stochastic-deterministic GO methods which could reduce computation time by one order of magnitude while guaranteeing robustness. Our goal here was to further reduce the computational effort without loosing robustness. We have developed a new procedure based on the scatter search methodology for nonlinear optimization of dynamic models of arbitrary (or even unknown) structure (i.e. black-box models). In this contribution, we describe and apply this novel metaheuristic, inspired by recent developments in the field of operations research, to a set of complex identification problems and we make a critical comparison with respect to the previous (above mentioned) successful methods. Robust and efficient methods for parameter estimation are of key importance in systems biology and related areas. The new metaheuristic presented in this paper aims to ensure the proper solution of these problems by adopting a global optimization approach, while keeping the computational effort under reasonable values. This new metaheuristic was applied to a set of three challenging parameter estimation problems of nonlinear dynamic biological systems, outperforming very significantly all the methods previously used for these benchmark problems.