No n_reactions_thermo in TFA results when building thermodb from cobra model

[ElderMedic]

No description provided.

[ElderMedic]

Hey, I tried to run your tutorial script in dockerized pytfa environment: https://github.com/EPFL-LCSB/pytfa/blob/master/tutorials/tutorial_equilibrator.py And trying to build thermo db based on cobra model glycolysis.mat and also the reduced e-coli model.

As the result shows I have 0 reaction in the model that have thermodynamic constraints applied to them.

`
2024-11-13 16:55:56,318 - thermomodel_new - INFO - # Model preparation done.
2024-11-13 16:55:56,319 - thermomodel_new - INFO - # Model conversion starting...
2024-11-13 16:55:57,290 - thermomodel_new - INFO - # Model conversion done.
2024-11-13 16:55:57,292 - thermomodel_new - INFO - # Updating cobra_model variables...
2024-11-13 16:55:57,299 - thermomodel_new - INFO - # cobra_model variables are up-to-date
value
key
name new
description new
num constraints 1019
num variables 1287
num metabolites 215
num reactions 268
value
key
num metabolites(thermo) 215
num reactions(thermo) 0
pct metabolites(thermo) 100
pct reactions(thermo) 0
Some fluxes
DM_4CRSOL 0.000118
DM_5DRIB 0.000122
DM_AMOB 0.000001
DM_MTHTHF 0.000709
Ec_biomass_iJO1366_WT_53p95M 0.529469
Name: fluxes, dtype: float64

Onjective solution: 0.5294692065760227

Total sum of fluxes: 160.31493578399147
`

Is this expected? Does it mean the result/model is not reliable/feasible?

[ElderMedic]

I also have many "2024-11-13 17:39:56,907 - eQuilibrator_formation - DEBUG - 248 : thermo data NOT created, error : 'ComponentContribution' object has no attribute 'dG0_prime'" such error in the equilibrator logs.

[ElderMedic]

def compute_dGf(compound, cc):
    """Get ΔGf from equilibrator `compound`."""
    dG0_prime, dG0_uncertainty = cc.dG0_prime(
        PhasedReaction(sparse={compound: 1}, rid="tmp_"+compound.id)
    )
    return dG0_prime, dG0_uncertainty

I don't find cc.dG0_prime() method in equilibrator_api.ComponentContribution, so consider this a bug.
Change it to standard_dg_prime instead does not immediately fix everything and got me a lot of new issues, so the input is not correct now...

[ElderMedic]

and why is charge_std=None in the returned dict of equilibrator.compound_to_entry?

> /src/pytfa/pytfa/thermo/utils.py(67)check_reaction_balance()
-> sum_charge += metCharge * metCoeff
(Pdb) p metCharge
None

This will no doubt cause following process to fail...

[ElderMedic]

Please dear dev, share your docker container that you did all your testing. I can't even build and run your dockerfile without hardcoding each package's version.