/eMASS

Primary LanguageMathematica

eMASS

Here you can find all code used in chapter 3 of my PhD thesis together with respective data.

Overview

mathematica_code

Contains some Mathematica code used in Mathematica notebooks.

data

Contains omics data and equilibrium constants.

eMASSpy

Contains all Python code used to analyse and plot results from the models generated through eMASS and eMASS2.

enzyme_level_models-code

Contains the Mathematica notebooks used to build each enzyme-level model.

enzyme_level_models-data

Contains the data generated by building each enzyme-level model.

old_model_building_flux

Contains all code to build system-level models using eMASS.

plots

Contains all the plots in the thesis, chapter 3.

The code has only been tested on Linux Mint 17.3. To build the enzyme-level kinetic models we used Mathematica 11.0 and the following packages

To analyse the results, both Mathematica and the code in eMASSpy were used. The package requirements for the latter can be found under eMASSpy.

Reproducing results

To reproduce figure 3.4 in the thesis:

Run plot_eigenvalues.py under eMASS/eMASSpy/src/process_results. You'll need to change base_dir. The plot will be stored in eMASS/plots.

To reproduce figure 3.5 in the thesis:

Run plot_free_met_prediction.py under eMASS/eMASSpy/src/process_results. You'll need to change base_dir. The plot will be stored in eMASS/plots.

To reproduce figure 3.6 in the thesis:

Run plot_omics_dist.py under eMASS/eMASSpy/src/process_results. You'll need to change base_dir. The plot will be stored in eMASS/plots.

To reproduce figure 3.7 in the thesis:

Run plot_enz_prediction.py under eMASS/eMASSpy/src/process_results. You'll need to change base_dir. The plot will be stored in eMASS/plots.

To reproduce figure 3.8 in the thesis:

Run plot_time_courses.py under eMASS/eMASSpy/src/process_results - might take a while. You'll need to change base_dir. The plot will be stored in eMASS/plots.

To reproduce the data behind the figures from scratch

First you need to generate the enzyme-level models, in particular get different rate constant sets for each enzyme. These are then used by eMASS and eMASS2.

Enzyme-level models

  1. Go to eMASS/enzyme_lelve_models-code and run each notebook, in each notebooks you will need to change:
  • "pathMASSef" - path to where you have MASSef, so that it can find the data Under "Simulate data", run only the subsection "Simulate data without uncertainty"
  1. Open "gather_enz_data.sh" change the path to eMASS in the respective variables and run it to copy the results to the appropriate folders.
System-level models generated through eMASS2
  1. Open model_prototype_8enz.nb, change "mainDir" and run the sections
    • "Initialize notebook"
    • "Model generation" This will create all the data needed to fit the free metabolite and total enzyme concentrations and store it in eMASSpy/data_fitLabel.
  2. On an HPC cluster create a folder that will contain all data and results, e.g. concentrations_fit;
  3. inside the folder "concentrations_fit", create two folders: "data" and "results";
  4. copy the data on eMASSpy/data_8enz to the HPC cluster: concentrations_fit/data;
  5. fit free metabolite and total enzyme concentrations using Python in an HPC cluster. 4.1 copy the contents of eMASSpy/run_on_cluster to concentrations_fit; 4.2 use prep_bash_scripy.pt under eMASSpy/run_on_cluster to set the jobs running. Here you need to change base_dir, and you probably need to change some parameters in prep_bash_script().
  6. once all the jobs have ran, copy the results in concentrations_fit/results to eMASSpy/cluster_results/results_fitLabel, also copy eMASSpy/data_fitLabel
  7. analyze the results by running analyze_results_eMASS2 under eMASSpy/process_results
  8. go back to model_prototype_8enz.nb, here you need to run again
    • "Initialize notebook"
    • "Model generation" - except "Export flux equations" and "Export rate constants"
  9. In "Model building", run:
    • "Plug-in concentrations from python"
    • "Build and export models" or "Import models"
    • "Get models with different initial ATP concentrations"
    • "Select subsets of models with low/high and in between ATP concentrations"
    • "Export subsets of models with low/high and in between ATP concentrations"
  10. Run everything under "Model simulation" except possibly "Plot simulations"
  11. Run everything under "Model analysis"
System-level models generated through eMASS

  1. Go to eMASS/old_model_building_flux, and run all enzyme_name.nb notebooks - you will need to change "baseDir" in each notebook though.

  2. Open whole_model_building.nb, change "baseDir", and run:

    • "Initialize notebook"
    • "Build models"
    • "Model simulation"
    • "Model analysis"

Contact: marta.ra.matos@gmail.com