A Systematic Modeling Framework for Biochemical and Biological Systems

In: 7th International Conference On Systems Biology, Yokohama, Japan. October, 2006

Authors

• Henning Schmidt
• Mats Jirstrand
• Gunnar Cedersund

Abstract

Dynamic modeling of biochemical and biological systems is an intensively interdisciplinary process. Tasks to be performed are the generation of quantitative data by experiments, setting up biological and mathematical hypotheses about the considered processes, estimation of model parameters from the measured data, and computational analysis of the resulting models. In engineering, so called, system identification techniques (i.e. building dynamic models from measurement data) have been developed and successfully applied over the last 30 years and a large toolbox of system theoretic results and computational tools is available. However, these can not directly be applied to the modeling of biological systems. The main reason lies in different desired levels of abstraction. In engineering applications the focus is mostly on the prediction of the systems input/output behavior, while the purpose of models of biological systems is to enhance the understanding on how the components within a certain system interact in order to create the functionality of the system. In this work we present a systematic modeling framework for biochemical and biological systems. We focus on dynamic models based on ordinary differential equations; however, the proposed approach is applicable as well to systems requiring alternative mathematical formulations, such as partial differential equations. The main steps of the proposed modeling framework are: a) definition of a model network structure, b) determination of mathematical rate expressions, c) parameter determination (experiment design, identifiability analysis, parameter estimation, model reduction), and d) model validation. The framework includes the fact that the modeling process will, in general, not be sequential but that the different steps need to be repeated in a problem dependent cyclic manner. The goal of the presented modeling framework is on one hand to give a guideline to modelers, but most importantly it serves to highlight parts of the modeling process for which important system theoretic results and/or computational tools are strongly needed. A minor result of the framework is the insight that computational tools used for different steps need to focus on the needs of the users that are supposed to use them. This might seem a trivial result; however, within systems biology presently available computational tools are often very difficult to use by the targeted user group and/or impose limitations that are not acceptable. We illustrate the use of the modeling framework by applying it to the development of a dynamic model for the complex interplay between neuron and glial cells.

BibTeX

 
@InProceedings{ Schmidt.Jirstrand.Cedersund, 
   title = { A Systematic Modeling Framework for Biochemical and Biological Systems }, 
   author = { Henning Schmidt and Mats Jirstrand and Gunnar Cedersund },   
   booktitle = { 7th International Conference On Systems Biology, Yokohama, Japan },             
   month = oct, 
   year = 2006, 
} 

---

This publication belongs to the project NeuronGlial.