How molecular should your molecular model be? On the level of molecular detail required to simulate biological networks in systems and synthetic biology

• Published in: Methods in enzymology Vol. 487; p. 171
• Main Authors: Gonze, Didier, Abou-Jaoudé, Wassim, Ouattara, Djomangan Adama, Halloy, José
• Format: Journal Article
• Language: English
• Published: United States 2011
• Subjects: Genes, Regulator; Models, Biological; Models, Molecular; Systems Biology - methods
• ISSN: 1557-7988
• DOI: 10.1016/B978-0-12-381270-4.00007-X

The recent advance of genetic studies and the rapid accumulation of molecular data, together with the increasing performance of computers, led researchers to design more and more detailed mathematical models of biological systems. Many modeling approaches rely on ordinary differential equations (ODE) which are based on standard enzyme kinetics. Michaelis-Menten and Hill functions are indeed commonly used in dynamical models in systems and synthetic biology because they provide the necessary nonlinearity to make the dynamics nontrivial (i.e., limit-cycle oscillations or multistability). For most of the systems modeled, the actual molecular mechanism is unknown, and the enzyme equations should be regarded as phenomenological. In this chapter, we discuss the validity and accuracy of these approximations. In particular, we focus on the validity of the Michaelis-Menten function for open systems and on the use of Hill kinetics to describe transcription rates of regulated genes. Our discussion is illustrated by numerical simulations of prototype systems, including the Repressilator (a genetic oscillator) and the Toggle Switch model (a bistable system). We systematically compare the results obtained with the compact version (based on Michaelis-Menten and Hill functions) with its corresponding developed versions (based on "elementary" reaction steps and mass action laws). We also discuss the use of compact approaches to perform stochastic simulations (Gillespie algorithm). On the basis of these results, we argue that using compact models is suitable to model qualitatively biological systems.