— the global portal for all things SBML

Ad-hoc features requested pre-2004

Last modified: 17 Dec. 2003 except where noted

This document lists features requested for SBML that are either: a) not fully developed to warrent a proposal document, b) not 'big' enough to warrent their own proposal document or c) do not fit into any given working groups scope.

  • Problem: SBML has no way to relate the two species together when they represent the same biochemical entity in two different compartments. Put another way SBML 'species' are really 'pools' a real species can have any number of pools in different compartments. Solution: create a new optional element speciesType with at least id and name attributes and then add a new speciesType attribute to species. speciesType elements then represent real species and species elements then represent pools. All these new components would be optional. (Update by MH on 5 September 2008: species types and compartment types were added in SBML Level 2 Version 2.) Issues to consider: How do we relate identical reactions that occur in different compartments? Raised by: Fabien Campagne, Jeremy Zucker and others
  • Problem: It can be important for reproducing a given numerical solution to specify the dependant variable on an algebraic rule. This is not possible in SBML. Solution: add an optional attribute dependantVariable on the algebricRule element that contains a variable id. Raised by: Mark Vass and Herbert Sauro
  • Problem: Its not possible to define functions within a mathml element using the sbml mathml subset even though this is possible in mathml. Solution: add declare to subset and allow lambda elements at the top-level with declare elements. Issues: variable scope aspects need to be checked carefully. In most contexts this is going to be redundant i.e. can be covered by SBML function definitions. Raised by: Nick Allen
  • Problem: It is not possible in SBML to specify the molecular weight of a species. Solution: perhaps integrate CML into SBML
  • Problem: SBML requires common biochemical constants to be defined everytime they are used. Raised by: Herbert Sauro
  • Problem: It is not possible to create unit definitions from unit definitions. Solution: enable the kind attributes of unit elements to refer to unit defintions as well as the unitkind enumeration.
  • Problem: We need to represent bounds on reaction flux, initial conditions and parameter values. Raised by Wayne Rindone
  • Problem: Need to label species for discrete simulation in hybrid simulation environments: Solution: add discrete, boolean {use="optional" default="false"} attribute to species element Raised by: Chris Cox et al.; see their comments on support for stochastic modelling. (Update by MH on 5 September 2008: this was one of the reasons SBO was developed and supported in SBML.)

The following items are within the controlled vocabularies working group scope but are placed here for save keeping:

  • Problem: It is not possible to label algebraic rules that are conservation laws. Note: solution requires CV for rules. Raised by: Marc Vass and Herbert Sauro. (Update by MH on 5 September 2008: this can now be done using SBO annotations.)
  • Problem: more straight forward way to extract mass action kinetic rate parameter from rate laws. Solution: label reaction parameter using CV. Raised by: Chris Cox et al.; see their comments on support for stochastic modelling.

Retrieved from ""

This page was last modified 21:02, 5 September 2008.

Please use our issue tracking system for any questions or suggestions about this website. This page was last modified 21:02, 5 September 2008.