Parameter Estimation Using Divide-and-Conquer Methods for Differential Equation Models

[摘要] In systems biology, a key topic is the elucidation of the dynamicbehavior of biological processes that are made up of complexbiochemical networks. Statistical modeling is an important to capturethe dynamics of biochemical networks such as metabolic networks,signal transduction pathways, and gene regulatory networks. Нesebiochemical models have a set of parameters that represent thephysical properties of the systems, such as kinetic constants andreaction rates. In general, the development of these models requirestwo steps: model structure construction and parameter estimation. Нemodels are oіen constructed with time derivative expressions, such asordinary diوٴerential equations (ODEs), to describe the change ofcertain quantities of interest over time [1,2]. Нe model parameters arethen estimated by simulating the actual processes obtained fromexperimental analyses [3-5]. However, because the diوٴerentialequation model has many uncertain parameters and limitedmeasurement data, parameter estimation is a major bottleneck in thedevelopment of useful biochemical models [6,7].