Feat: add pytfa metabolomics

[carrascomj]

This pull request adds pytfa to simulations and an interface to access the thermodynamic flux balance analysis (TMFA). This also includes a new simulation method: tmfa. At the moment, the method can just be used if metabolomics data is provided, therefore the models are prepared only when TMFA is called, to avoid useless computation.

The equilibrator side failed for some of the models so I decided to go with the thermoDB provided in the pytfa repository.

Unresolved yet:

• Decide if "seed_id" (not "seed.compound") annotation should be added to to every model's metabolites or translate the thermoDB to CHEBI. I have already implemented and succesfully tested the latter, but includes a hack, since the "seed_id" annotation is hardcoded in pytfa.
• Manage growth input from the user. Should the relaxation be on the thermodynamics constraints (implemented in pytfa but needs some adjustments) or in the metabolic constraints?
  - Answer: both. If the model cannot grow, thermodynamics constrained are relaxed. After that, growth is checked again and further relaxations on the metabolomics constraints are applied if needed.

[codecov-io]

Codecov Report

❗ No coverage uploaded for pull request base (devel@b667057). Click here to learn what that means.
The diff coverage is 78.65%.

@@           Coverage Diff            @@
##             devel     #155   +/-   ##
========================================
  Coverage         ?   76.71%           
========================================
  Files            ?       23           
  Lines            ?     1078           
  Branches         ?        0           
========================================
  Hits             ?      827           
  Misses           ?      251           
  Partials         ?        0

Impacted Files	Coverage Δ
src/simulations/schemas.py	100% <100%> (ø)
src/simulations/storage.py	84.31% <100%> (ø)
src/simulations/resources.py	76.4% <35.71%> (ø)
src/simulations/modeling/simulations.py	90.47% <66.66%> (ø)
src/simulations/modeling/pytfa_helpers.py	84.74% <69.23%> (ø)
src/simulations/modeling/community.py	93.02% <97.36%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update b667057...20491bd. Read the comment docs.