bug: makeEcModel if cytosol is not 'c'

[edkerk]

As reported by @ckitti, if the model does not use 'c' to indicate cytosol, makeEcModel fails when adding e.g. protein_pool pseudometabolite. Compartment code should be a parameter in ModelAdapter.

[Yu-sysbio]

Sorry I am a bit confused here. In the reported case, what are there in model.comps and model.compNames?

[edkerk]

model.comps contains Cytosol, while in makeEcModel it assumes to be c. While the latter is more common, it will be little effort to allow for alternative compartment IDs.

GECKO/src/geckomat/change_model/makeEcModel.m

Line 321 in 6c0dae9

pool.compartments = 'c';

[Yu-sysbio]

I see. Then what is the following line used for?

GECKO/userData/yourModel/yourModelAdapter.m

Line 71 in 6c0dae9

obj.params.enzyme_comp = 'cytoplasm';

What I do in my case is to find the compartment ID of cytosol in model.compNames, which is Cytoplasm in my model, then I just set obj.params.enzyme_comp = 'Cytoplasm'. I never check what there are in model.comps. Is this correct?

Is obj.params.enzyme_comp strictly needed?

[edkerk]

It's a remnant of GECKO1&2's parameters function, it hadn't been used in the GECKCO3 code yet. But well spotted, I'll just refactor makeEcModel to consider obj.params.enzyme_comp = 'Cytoplasm' and you indeed do not need to care about model.comps abbreviations.

And it is needed to specify a compartment when adding the proteins and protein pool pseudometabolites. It would work perfectly fine it these would be added to any random compartment, but it would look odd, so best to ensure it is in cytosol.

[edkerk]

Fixed in 3f9d897