Developing next generation predictive models: a systems biology approach

摘要

In the area of predictive microbiology, most models focus on simplicity and general applicability, and can be classified as black box models with the main emphasis on the description of the macroscopic (population level) microbial behavior as a response to the environment. Their validity to describe pure cultures in simple, liquid media under moderate environmental conditions is widely illustrated and accepted. However, experiments have shown that extrapolation to more complex (realistic) systems is not allowed as such. In general, the applicability and reliability of existing models under more realistic conditions can definitely be improved by unraveling the underlying mechanisms and incorporating intracellular (microscopic) information. Following a systems biology approach, the link between the intracellular fluxes and the extracellular measurements is established by techniques of metabolic flux analysis. Flux balance analysis (FBA) uses an objective function to derive, through optimization over the solution space of this underdetermined linear system, an intracellular flux distribution. A first section of this paper discusses the background of the FBA approach. The modeling approach presented in this paper can lead in the future to more accurate predictive models for more complex systems, such as co-cultures and growth as colonies, based on a top-down systems biology approach. A second section focuses on an important issue with respect to the implementation of FBA analysis when modeling dynamics systems. To be able to use flux predictions by FBA in a dynamic model, all degrees of freedom of the underdetermined linear system need to be removed. In most cases, not all degrees of freedom can be fixed through optimization and multiple flux solutions are found. Extra information is required to identify the remaining degrees of freedom. A study on the minimum number of constraints (measurements) needed to remove all degrees of freedom is determined for a small-scale metabolic reaction network constructed for E. coli K12.