To hybrid QM/MD, to molecular mechanics models. The approach outlined in the paper can provide a detailed model of metabolism that provides in-depth information, but not all questions may require this level of information. Here, the basic aspects of the statistical thermodynamics background needed for simulating metabolic systems are presented. The methods section does require some mathematical background in multinomial statistics, however this background is not necessary to understand the application presented in the results section. The application is that of the tricarboxylic acid cycle from Escherichia coli, for which the free energy, energy and entropy profiles are determined as well as predictions of metabolite concentrations. However, the point of this report is not to model a particular process in high fidelity, but rather to demonstrate the principles of applying statistical thermodynamics to metabolic reaction networks. Finally, this report concludes with a discussion of the advantages and limitations of using state-based simulations to model metabolism. Unfortunately, it’s not currently possible to obtain all the necessary rate constants to model a system with specific time dependence. Besides the fact that each ortholog of an enzyme will have different rate constants, the challenge of obtaining accurate rate constants is much harder than one might imagine. INCB28060 1029712-80-8 kinetic parameters vary significantly with solution conditions – pH, ionic strength, dielectric, etc. While thermodynamic parameters also vary with solution conditions, the variation is significantly more predictable using modern computational chemistry methods. In fact, useful estimates of standard free energies of reaction can be obtained en mass for large scale modeling from resources such as the Thermodynamics of Enzyme-Catalyzed Reactions Database at NIST, the Biochemical Reactions Thermodynamics Database, and the eQuilibrator web server. Given the variability of kinetic parameters due to physical influences and differences in rates between orthologs, it is debatable whether achieving a full-scale kinetic simulation is a reachable goal. Currently, flux-based models are the best that one could do for modeling large-scale processes in metabolism. Fluxbased approaches are not based on law of mass action, so prediction of energy requirements and metabolite levels is difficult without assumptions regarding the relationship between flux and free energy changes. In this light, the development of metabolic models based on statistical thermodynamic.