PyFBA 1.2 release

Our genome-scale metabolic model software, PyFBA, has been upgraded to version 1.2 recently. With new features ranging from a new Model class to Python 3 support, this release expands the usefulness of PyFBA. Version 1.2 has been updated on our GitHub master branch and PyPI. Read further to see more details.

What’s new in v1.2?


  • PyFBA now works with Python 3! There were a few issues that we had to address regarding unicode characters, which you can read more about here or here. Basically, Python 3 replaced Python 2’s str type with Python 2’s unicode type, which it now calls str (yes, I know, it’s already confusing). Anyways, we got it working.


  • Obtain SEED complexes associated with a set of SEED roles using the function PyFBA.filters.roles_to_complexes().
  • A new step in the gap-fill algorithm has been implemented that searches for reactions associated with the EC numbers. The EC numbers must have been present in the SEED roles used to generate the model.
  • Introduction of the new Model class. This was developed in order to make typical model functions easier for users.
    1. Generate a Model object using a SEED roles file: PyFBA.model.roles_to_model().
    2. Run FBA with the Model object: my_model.run_fba().
    3. Save and load the Model to and from hard disk: PyFBA.model.save_model() and PyFBA.model.load_model().
    4. Gap-fill the Model with a Media file: my_model.gapfill().
    5. You can view how to use these functions in our previous blog posts:


  • New sets of iPython Notebooks were created to give users examples on how to use PyFBA. All Notebooks can be found on our GitHub page here.
    1. Find a metabolite
    2. PATRIC to FBA
    3. Saving and loading a model
    4. Gap-fill a model


  • During the gap-fill algorithm, some portions were taking longer than necessary to find the optimal number of reactions to add to the model to reconcile growth. These were fixed and patched up.

If you’re interested in learning more about the tool, read the published version of PyFBA version 1.0 here.