Automating Mathematical Modeling of Biochemical Reaction Networks

Abstract

In this chapter we introduce a five-step modeling pipeline that ultimately leads to a mathematical description of a biochemical reaction system. We discuss how to automate each individual step and how to put these steps together. First, we create a topology of interconversion processes and mutual influences between reactive species. The Systems Biology Markup Language (SBML) encodes the model in a computer-readable form and allows us to add semantic information to each component of the model. Second, from such an annotated network, the procedure known as SBMLsqueezer generates kinetic equations in a context-sensitive manner. The resulting model can then be combined with already existing models. Third, we estimate the values of all newly introduced parameters in each created rate law. This procedure requires that a time series of quantitative measurements of the reactive species within this system be available, because we calibrate the parameters with the aim that the model will fit these experimental data. Fourth, an experimental validation of the resulting model is advisable. Fifth, a model report is generated automatically to document the model with all of its components. For a better understanding, we will begin with an introduction to current standardization attempts in systems biology and generalized approaches for common rate equations before discussing computer-aided modeling, parameter estimation, and automatic report generation. We complete this chapter with a discussion of possible further improvements to our modeling pipeline.