Forums

In testing our own software, we've recently downloaded and imported about 20
of the models now available on BioModels.net. This is a tremendous resource
for us.



I noticed that some reactions in these imported models fail to include
listOfModifiers structures even when it is clear from the kineticLaw that
modifiers are intended. I reported one such problem to the curators of the
biomodels database, and Nicolas suggested I bring this topic to the
attention of the entire SBML community.



Clearly the absence of listOfModifiers will not alter the resulting
numerical solutions as long as the modifiers are included appropriately in
the kineticLaws, but some applications (ours for example) intend to flag
kineticLaws as having errors if they refer to species that are not included
in the listOfReactants, the listOfProducts or the listOfModifiers.
Furthermore, when an application produces a drawing of a model, it is likely
to do so based on the lists of reactants, products and modifiers, rather
than trying to parse the rate laws. Hence, models that omit listOfModifiers
will often be rendered inaccurately or at least incompletely.



The L2v1 spec does say that the listOfModifiers is optional, but one could
argue that in order to be fully L2v1 compliant, the listOfModifiers should
be required if the kinetic law refers to a species that is neither created
nor destroyed.



<informal poll>What are other developers doing? Does everyone write a
listOfModifiers for reactions that involve modifiers? Should appropriate
list(s)OfModifiers be required for L2v1 compliance?</informal poll>



Robert



Robert D Phair, PhD

Chief Scientific Officer

Integrative BioInformatics Inc

Rockville, MD 20854 USA

301.315.8114