Error Analysis during Estimation of Metabolic Fluxes through Metabolite Balancing

Abstract

Accurate estimation of fluxes in a bioreaction network is critical as metabolic fluxes are a key determinant of cell physiology and function. The input data for metabolic flux estimation by the metabolite balancing technique includes the growth rate of cells and specific nutrient uptake and metabolite production rates. These rates are computed from experimental measurements of the primary cell, nutrient and metabolite concentrations. As a result, any error in experimental measurements of the primary variables propagates into the rate estimates, which in turn affect the accuracy of the resulting metabolic fluxes. In this study, we present analysis of the error associated with computing metabolic fluxes using a combination of experimental and theoretical approaches. Baby Hamster Kidney (BHK) and Chinese Hamster Ovary (CHO) cells were cultivated in perfusion culture at approximately 20 × 106 cells/mL. Samples were periodically drawn from the bioreactors and analyzed for cell, nutrient and metabolite concentrations. The error associated with these primary variables was first determined by performing multiple measurements after which the Gaussian law of error propagation was used to estimate errors associated with specific rates and metabolic fluxes in an established metabolic network. This analysis helped establish relationships between the errors in metabolic fluxes and those in the primary variables allowing a priori prediction of the errors in metabolic fluxes.