This specification also does not support several features of interest including: compartmentalization of entities, the transport of entities, and representation of stoichiometric coefficients.
Stimulation and Necessary Stimulation do not include transporter and channel activities, at least not in the official 2006 or 2010 specs. I guess they extended it as a
matter of convention, just like we conventionally extended catalysis. Frankly, neither are correct, but I’m fine with adopting the same convention as Augustin, et al. if it helps interoperability, etc.
However, my only argument against this would be that if we’re going to go through the trouble of finding all the “wrong” catalysis interactions to switch them to necessary stimulation, we’d be better off switching them to a new, unique interaction type, i.e. a placeholder for actual transport activity. Otherwise, we’ll be in this same spot later on down the road when we have to find all the “wrong” necessary stimulations and switch them to the eventual correct one (which does not exist in any official standard today).
But other that this point, I’m fine with it. It’s not more wrong than what we’ve got now :)