Fig. 2

Model validation. Model predictions compared to experimentally measured data (blue). All model adjustments and conversions are described in Additional File 2: Text S2. The bars represent the mean of the different data, while error bars indicate the range of the data (minimum and maximum). A Simulated NADH production flux, stoichiometrically converted to O₂ consumption, compared to the uncoupled oxygen consumption flux of permeabilised MCAD-KO and control HepG2 cells. In converting an NADH production flux to an oxygen consumption flux, the contribution of FADH₂ and reducing equivalents from the downstream TCA cycle were also considered (Additional File 2: Text S2). The experiment and simulation contained 25 µM of palmitoyl-CoA and 2 mM of L-carnitine (blue; n = 4 in both groups). B Simulated acetyl-CoA production flux, stoichiometrically converted to a ketogenic flux, compared to measured ketogenesis in 24 h-fasted, healthy human subjects (n = 17). C, D Comparison of simulated and measured palmitoyl-CoA (C16-CoA, C) and octanoyl-CoA (C8-CoA, D) dehydrogenation rates (acyl-CoA substrate at 30 µM) in crude lysate compared to cell lysates after immunoprecipitation of VLCAD or MCAD, respectively (blue; n = 3 or n = 4)