Recent Developments in the Systems Biology Markup Language

Michael Hucka1, Andrew Finney2, Benjamin Bornstein, Bruce Shapiro, John Doyle, Hiroaki Kitano, Caltech;, University of Hertfordshire

Over the years, a variety of biochemical network modeling packages have been developed and used by researchers interested in understanding cellular networks. No single software package currently answers all the needs of the biology community; nor is one likely to do so in the near future, because the range of tools needed is vast and new techniques are emerging rapidly. It seems unavoidable that, for the foreseeable future, systems biology researchers will continue using multiple software packages to carry out their work. This diversity of resources results in several problems, including difficulties in moving models between packages and a lack of a straightforward mechanism for researchers to publish models in electronic form.

In response to these needs, the Caltech ERATO Kitano group developed the Systems Biology Markup Language (SBML) in collaboration with a community of software developers and computational modelers. SBML is a machine-readable model definition language based upon XML, which in turn is a simple and portable text-based medium that has been gaining widespread acceptance in computational biology and bioinformatics. SBML's current vocabulary consists of structures for specifying fundamental aspects of a system of biochemical reactions, in particular the chemical species involved in the reactions, the reaction kinetics, the volumetric compartments in which the reactions take place, definitions of mathematical functions, parameters in the system, units on quantities, discrete events and delays, and additional mathematical constraints.

The feature set for SBML Level 2 (the latest version of SBML) has recently been settled, and a final specification will be issued later this year. Work has already begun on SBML Level 3. In this poster, we will provide an overview of SBML with an emphasis on the main changes introduced in Level 2. These changes include: replacing SBML Level 1's text-string based format for mathematical expressions with MathML, introducing support for metadata using the same metadata as CellML, introducing support for named function definitions, and introducing new constructs for discrete events and time delays. We will also describe two SBML parser libraries now available under the terms of the LGPL open-source license: libsbml, a parser/writer library written in C and intended to be embedded in applications, and MathSBML, a package written in Mathematica that provides the ability to translate SBML models to systems of ODEs as well as solve and plot the results. Finally, we will briefly discuss some of the features currently being developed for SBML Level 3. These include the ability to compose a model out of submodels and the ability to have arrays of elements.

Software packages that support SBML include: BASIS (University of Newcastle), Cellerator (JPL), Cytoscape (ISB), Dizzy (ISB), Gepasi (Virginia Tech.), Jarnac (Keck Graduate Institute), JDesigner (Keck Graduate Institute), JigCell (Virginia Tech.), BioSketchPad (BBN), NetBuilder (University of Hertfordshire), SBedit (ERATO Kitano), SBW (Caltech \& University of Hertfordshire), StochSim (Cambridge University), and Virtual Cell (UCHC). SBML is also the de facto model interchange standard used by the DARPA BioSPICE and IECA consortia.

More information about SBML is available online at