Fig. 5.

A Bayesian approach to determine important metabolites or mass isotopomers for estimating metabolic flux levels. Computer simulations were performed using the expanded toy model for the case of 50% [U–¹³C]glucose as the carbon source. (a and b) Posterior distributions of metabolic flux level of v₄ determined using Metropolis–Hastings method when +3 isotopomer (a) and +0 isotopomer (b) of OAC were used as the measured data. (c) Comparison of 95% intervals of posterior distributions when each isotopomer, i.e., OAC, AKG, and PEP, was used as the measured data. In all simulations, [U–¹³C]glucose and non-labeled glucose at a 50:50 ratio were used as carbon sources.