Data files

Comparison of model SS metabolite concentrations with measured values using 1000x sampling from the Gausian distribution of the measured values based on multiple replicates per measured conditions. Graphs showing the distribution of measured and simulated metabolite concentration for 95 mutand (KO, OE), perturbation and time series measurements. Model simulations performed using 24h proteomics with modification of enzyme parameters for KO and OE mutants.

Contains a Jupyter notebook file that uses libroadrunner and tellurium to run all simmulations and analysis based on the 40 independent samples. The Readme.txt file contains information on how to recreate the complete modelling environment used for all simmulations and analysis using Anaconda.

Local sensitivity analysis based on 40 samples using 1000x sampling from measurement distribution. The control shown is the control over flux through glycolysis represented by flux through PRK. Th plot summarized the control for each parameter over all observed metabolite concentrations encountered for the 40 samples. As such the metabolic control analysis is local but shows the distribution taking into account measurement error as well as biological variation over the 40 samples.

Tab seperated file containing the raw output of the local sensitivity analysis based on 40 samples (based on 1000x sampled metabolite values) from the MEAN and SD of the metabolite measurements. Sensitivity analysis is based on the flux through PFK as objective and as proxy for flux through glycolysis. Data can be plotted using the R script "plotLocalGlobalSensitivity1.5.R" associated to the same assay.

Contains relative mutant (OE, KO) perturbation and time series samples metabolite concentrations and enzyme fold change of targeted enzymes used for model validation. Measured are the relative fold change, Mean and SD of log2 fold change values are based on multiple measurements per sample (minimum of three). Contains input data for Automated Model simulations pipeline to load and update the models metabolite concentrations and enzyme parameters to simulate all sample using a custom python script ...

Symmetric mean absolute percentage error per sample graph for the 40 independent samples

Contains relative metabolite concentrations for 40 samples based on technical triplicates. Medium and SD values were calculated and used for 1000 sampled simmulations (sampling from the measurement distribution per metabolite) per sample. Also contains annotion to link metabolite concentrations and protein fold change measurements for OE and KO mutants to the model as well as external glucose, acetate and lactate concentrations. A SBtab like format was used to easily load the MEAN and SD metabolite ...

Violin plot of the metabolic control of model parameters on the flux through PFK (as a proxy for flux through glycolysis) based on a 100.000 Latin Hypercube samples from the parameter space (range 0.001-100 for Km values and 0.001-1000 for Vmax values).

Shows the correlation in metabolic control between parameters. The plot shows central carbon metabolis basically consist of a few control hubs of reactions of which the parameters are correlated. In other words, just the reaction network combined with allosteric control and equilibrium constants impose some constrains on possible parameter value combinations that lead to certain behaviour (such as the flux and concentrations measured in vivo/vitro).

Contains the raw output of the global sensitivity analysis, can be used as input for plotting using the R script "plotLocalGlobalSensitivity1.5.R" associated to the same asset