Forums

The topic of this vote is:
INTRODUCING COMPARTMENTTYPES

The URL for the voting page is:
http://www.surveymonkey.com/s.asp?u=241921711592

Additional background is available in the archive of sbml-discuss at
http://www.sbml.org/forums/index.php?t=tree&goto=2714&rid=0

The SBML Level 2 Version 2 Specification is at:
http://sbml.org/wiki/sbml-level-2-version-2.pdf

The draft specification for SBML Level 2 Version 2
introduces a new object called "SpeciesType". (See section
4.6 in the specification document noted above). A
SpeciesTypes refer to a species independent of its location,
and was introduced to permit relating multiple species of
the same type together. In a sense, speciesType represent
"classes" of interacting partners, and SBML species are
"instances" of speciesType. SpeciesTypes give the identity
of the interacting partners, whereas SBML species objects
represent physical populations having such properties as
concentration or amount. Such a distinction is natural and
exists in the data model of most pathway databases.

Nicolas Le Novere recently argued that a similar situation
applies to compartments, and with the advent of SpeciesTypes
in SBML, the addition of a "compartment type" concept would
make sense and add consistency to the SBML data model.

The idea of the CompartmentType is to relate several
compartments of the same type together, analogously to how
SpeciesType works. A CompartmentType could be used to
define the generic properties of compartments, while SBML
"compartments" would be the actual instances, representing
containers with quantified properties such as size. Some
examples of compartment types include:

* All mitochondria in a cell;
* All compartments having certain reactions, such as ADP+P->ATP;
* All cells in a plant meristem;
* All voxels of the same type in a spatial model.

Similarly to SpeciesType, the new element CompartmentType
would be optional. SBML's Compartment element would have a
new attribute compartmentType, also optional. The final
impact on existing software systems would be minimal,
and all SBML Level 2 Version 1 models would remain valid
SBML Level 2 Version 2 models.

The original aim of SpeciesType was to allow the definition
of generalized reactions that are "identical" (same reacting
partners, same KineticLaw, same parameter values) in
different compartments. (See Andrew Finney's proposal at
http://www.cds.caltech.edu/~afinney/multi-component-species.pdf)
The existence of generalized reactions requires the
existence of CompartmentTypes, because the KineticLaw of a
generalized reaction will need to refer the symbol of the
generic compartment. However, generalized reactions will be
the topic of a different vote. Adopting CompartmentType now
does not trigger a requirement to adopt generalized
reactions.

Here are alternative for addressing CompartmentType for
SBML. Which do alternative do you believe should be adopted?

Option (a): Yes, compartmentType should become part of
SBML Level 2 Version 2.

Option (b): Although CompartmentType is an interesting
concept that should be discussed more in-depth, it should
be postponed to SBML Level 3. Do not adopt it for SBML
Level 2 Version 2.

Option (c): No, I do not think compartmentType is a
good idea for SBML, and I will explain why on the
sbml-discuss mailing list.

Option (d): I have no opinion in this matter.

Please cast your vote for one of the alternatives above by
using the voting page at

http://www.surveymonkey.com/s.asp?u=241921711592

Mike