Re: [Biomodels-net-discuss] [SBO] What if SBML speciesType and species disagree?

[Michael H. <mh...@ca...>]

  mhucka> I'm wondering if the following situation can occur:
  mhucka> suppose you have
  mhucka>
  mhucka> <model>
  mhucka> ...
  mhucka> <listOfSpeciesTypes>
  mhucka>   <speciesType id="X" sboTerm="Y" />
  mhucka> </listOfSpeciesTypes>  
  mhucka> ...
  mhucka> <listOfSpecies>
  mhucka>   <species id="label" speciesType="X" sboTerm="Z" />
  mhucka> </listOfSpecies>
  mhucka> ...
  mhucka> </model>
  mhucka>
  mhucka> where Y != Z.  Can this happen?  Should it be
  mhucka> considered an error?  If so, what kind of error?

  lenov> It is semantically problematic, but not obligatory
  lenov> an error. Y could be a parent of Z. 

You're right.  (Actually, I even realized this after sending
the message, but didn't correct myself.)

  lenov> The other issue we face is that a given
  lenov> species/speciesType can be characterised by
  lenov> different kinds of terms, following the wish of the
  lenov> author.
  lenov> 
  lenov> Y = macro-molecule
  lenov> Z = simple chemical
  lenov> 
  lenov> is an error
  lenov> 
  lenov> y = macro-molecule
  lenov> z = enzyme
  lenov> 
  lenov> is not ...

OK.  I think the bottom line is that no connection must be
require between the speciestype SBO term and the species'
SBO term, correct?

MH