samseaver
commented
12 years ago My current approach to the problem of creating and switching generic biomasses for the same model is to simply create bare-bone model files with no compounds and a single biomass reaction. These can then be printed via ./mdl-printmodel.
The following models are named after their biomass components, and the resulting biomass file can be used to test specific cases:
NucleotidesBiomass.45632 AminoAcidsBiomass.45632 BVitaminsBiomass.45632
If I create my own biomass file, with my own biomass equation, and use ./mdl-loadbiomass to load that equation for a particular model (from file), then the equation gets added to the BIOMASS table, but it does not get added to the model's local provenance.
If I were to include the same biomass equation in a makeshift model which I then import, and I then print out the same biomass file (which now has its own id in ModelSEED), I can then use ./mdl-loadbiomass to load this new file. So it seems that for a biomass equation to be added to a model, it must already practically exist in ModelSEED.
To compound things further, I cannot simply use the id of a biomass already in ModelSEED to load it into a model, I have to print out the biomass to file (via ./mdl-printmodel) and use that.
So, effectively, this issue boils down to two, well, three, OK, four things:
1) We need a ./mdl-createbiomass that will load from file -- it probably should generate a biomass id rather than let the user dictate it.
2) We need a ./mdl-loadbiomass that can simply take the id of a biomass already in ModelSEED.
3) We need a ./mdl-printbiomass
4) We could do with a ./mdl-findbiomass function that will take a comma-separated string of compounds, and find all the biomass equations that have all the compounds queried.