Synthetic biology design principles enable efficient bioproduction of Heparosan with low polydispersion index for the biomedical industry

This is the repository of the code we used to generate the results in "Synthetic biology design principles enable efficient bioproduction of Heparosan with low polydispersion index for the biomedical industry" manuscript.

Matlab Version

We used Matlab version 23.2.0.2485118 (R2023b) Update 6, we did not test if this code works with other Matlab versions.

Repository structure

Included within this directory are all the scripts required to execute the simulations and optimizations, yielding the results showcased in the manuscript. This repository encapsulates everything essential for reproducing the findings detailed in the paper.

/optimization

ODE Model

The file model_hep6p.m contains the ODE model definition. The parameters of the model are implemented in the function parameters.m.

Files to use MEIGO optimization algorithm

• File CostFunction_HEP.m contains the implementation of the cost function for our multiobjective optimization problem.
• The file optimize_with_MEIGO.m is used to call the optimization algorithm used MEIGO. Check the EXAMPLES folder within the MEIGO repository for more info on how to use it. In this case, the file optimize_with_MEIGO.m includes all the parameters necessary for the execution with our cost function CostFunction_HEP.m.

/robustness

Scripts for the robustness analysis.

• Simulation_and_plot_perturbation.m script is used to get the different plots for the manuscript to illustrate the results.
• Optimal_Values_Calculation.mlx script calculates the uncertainties of the optimum values using the gaussian process.
• distributuions_optim.mlx and distributuions_random.mlx scripts calculate the optimum and random distributions using values of PDI and Mw for the two sets of data: optimal values from the optimization and random values.
• level_diagram_plot.m plot the PDI data against MW, and the precursors concentraitons for Figure 4 of the manuscript.
• results_cL1_10_cL2_10_cU1_20_cU2_20_date_20240331_154057.mat has the optimization results in Matlab format.
• tablaOutput.mat compiles the results in a table format.

/fba

We include all the code used to develop the flux balance analysis in different files:

• The original metabolic SBML model iDK1463, written as a MATLAB object nissle_core.mat.
• The new metabolic SBML model including the heparosan biosynthesis pathway (metabolites, enzymes, and reactions), written as matlab object nissle_core_heparosan.mat.
• The COBRA file FBA_heparosan.mlx (Matlab LiveScript) to add the heparosan biochemical reactions, and to perform the FBA.