Applying the ETFL framework to the CHO genome-scale metabolic model.
This code requires pyTFA and ETFL to function properly. To begin, clone a fork of these repos as well as a fork of this repo to a local folder with the following structure:
└───src_folder
└───pytfa
└───etfl
└───choetfl
It is recommended that you not clone the main branch of your personal forks, but to create and clone an unstable development branch of these forks. You can then preserve the main branch of your fork as a synchronisation branch to keep up with any updates made to these packages.
Next, navigate to choetfl and create the conda virtual environment using the requirements.yml file provided with this code. This step requires having Anaconda installed on your machine, see installation instructions. From the src_folder:
cd choetfl
conda env create -f requirements.yml
conda activate choetfl_env
Next, navigate to pyTFA and install it as a package. Repeat this step for ETFL.
cd ../pytfa
pip install -e .
cd ../etfl
pip install -e .
Finally, an MILP solver such as Gurobi or CPLEX is required. Follow relevant installation instructions for the Python installation of your solver of choice. For example, if using Gurobi:
conda install gurobi::gurobi
You can run the test_installation.py script to ensure all libraries have successfully been installed with the correct version. You should expect the following output:
Testing library versions:
Python - 3.8.20
NumPy - 1.19.5
Pandas - 1.4.4
Cobra - 0.24.0
PyTFA - 0.9.4
ETFL - 0.0.2
GurobiPy - 11.0.3 (This will vary depending on the installed solver)