We present multiTFA, a multivariate thermodynamics-based metabolic flux analysis package for Python. The framework takes advantage of the reactions' Gibbs free energy covariance matrix to tightly constrain metabolic models using thermodynamic constraints. It represents an improvement over a previous thermodynamic metabolic flux analysis (tMFA) method described in [1].

This implementation requires a COBRA model as input, as well as additional compartment conditions and metabolite concentrations. It allows user to perform various thermodynamic analyses on COBRA models, such as thermodynamic metabolic flux analysis, variability analysis, or flux sampling. Please see below for further details.

To install multitfa from PyPI is as simple as: (We recommend using a virtual environment).

pip install multitfa

To install from source,

git clone https://github.com/biosustain/multitfa.git
cd multitfa
pip install .

We highly recommend the installing CPLEX. Although we support GUROBI solver, we noticed that it is slower and often stuck when solving quadratic constraint problems.

Please note, Installation takes upto 3 GB. This is to accomodate equilibrator-api database files.

Installation requires

• cobra
• depinfo
• optlang<1.4.6
• numpy
• scipy
• pandas
• equilibrator-api==0.3.2b7
• component-contribution==0.3.2b4
• equilibrator-cache==0.3.2b2

for testing purposes,

• pytest
• tox

An example test case was setup to get flux ranges of example script paper_data_example.py using tox. To run the test case, from inside multitfa directory

pip install tox
tox -e examples

The documentation is available online at readthedocs.

• Copyright © 2018, Novo Nordisk Foundation Center for Biosustainability.
• Free software distributed under the Apache Software License 2.0.

If you use multitfa in a scientific publication, please cite doi:10.1101/2020.12.01.407387.