Open pgaudet opened 5 years ago
From @ValWood on February 28, 2017 11:48
Super!
(The small-molecule metabolism list is really useful for metabolic modellers etc)
At the ontology call we discussed that 'small molecule' was difficult to define. A suggestion is to use 'metabolite' -> 'Catalytic activity, acting to interconvert metabolites' Is that any better ?
"metabolite" is what we are trying to capture with this term. However this would translate to "metabolite metabolic process", it looks weird but I guess it's clear enough.
Discussed with @amorgat - she has no objection to using 'metabolite'.
If there are no objections from the ontology group, I will go ahead and implement that.
Thanks, Pascale
Discussed with @thomaspd Plan is to
We'd like to have a script to do this at each GO release (ie including updated Rhea).
@pgaudet @thomaspd If I understand your concern correctly, I think that you don't have to query ChEBI ontology for that task. Rhea reactions involving macromolecules have participants of the type GENERIC (GenericPolypeptide / GenericPolynucleotide) or POLYMER. Reactions with ALL participants of the type CHEBI can be classified as a 'catalytic activity, acting to interconvert metabolites'.
Great ! That's very useful.
We now need to define a strategy.
@pgaudet where can I find documentation on GO release cycle / procedure?
@pgaudet thanks!
From @dosumis on February 9, 2017 16:12
Intention is to distinguish catalytic processes acting on an 'information biomacromolecule' (CHEBI:33695) from those acting on other molecules. We can't do this logically with the way ChEBI is built (we don't have a class representing the complement of 'information biomacromolecule' (CHEBI:33695). But we could derive a small molecule list using a closed world intepretation of ChEBI (e.g. via graph reasoning). @cmungall - Can you script adding a derived class like this to GOCHE?
_Copied from original issue: geneontology/molecular_functionrefactoring#35