SBML

SBML
Name	SBML
Developer	Systems Biology Markup Language Community
Initial release	2000
Latest release	See Levels and Versions
Written in	XML
Operating system	Cross-platform
License	Community-driven

SBML is a computer-readable format for representing computational models of biochemical reaction networks. It enables exchange among software like MATLAB, COPASI, CellDesigner, Tellurium, and JWS Online, and supports modeling paradigms used by groups at MIT, Stanford University, Harvard University, European Bioinformatics Institute, and National Institutes of Health. SBML facilitates reproducibility in projects such as Human Cell Atlas, ENCODE Project, Cancer Genome Atlas, Physiome Project, and BioModels Database.

Overview

SBML provides an XML-based syntax that encodes entities such as species, reactions, compartments, parameters, and events so that tools including Systems Biology Workbench, SBMLtoolbox, LibSBML, COPASI, and CellDesigner can interoperate. The format is widely used by consortia at European Molecular Biology Laboratory, Max Planck Society, Wellcome Trust Sanger Institute, Cold Spring Harbor Laboratory, and Los Alamos National Laboratory to share models underpinning studies like AIDS research, Alzheimer's disease research, Type 2 diabetes research, cardiac electrophysiology, and metabolic engineering. Because the specification is machine-readable, teams using Google DeepMind, IBM Research, Microsoft Research, Amazon Web Services, and NVIDIA can integrate SBML into pipelines for simulation, parameter estimation, and sensitivity analysis.

History and Development

The SBML initiative emerged from meetings involving researchers from California Institute of Technology, Massachusetts Institute of Technology, European Bioinformatics Institute, University of Cambridge, and University of Oxford in the late 1990s and early 2000s. Early adopters included developers of COPASI, CellDesigner, Jarnac, BioSPICE, and SBW who sought interoperability like that achieved by standards such as XML and by projects like Gene Ontology and UniProt. Over time governance evolved with input from institutions such as National Science Foundation, Wellcome Trust, European Commission, and National Institutes of Health and with participation by laboratories at Princeton University, Yale University, University of California, Berkeley, and University of Tokyo.

Structure and Core Concepts

SBML models consist of hierarchies: model metadata, lists of species, compartments, parameters, reactions, rules, and events. Implementations often rely on libraries like LibSBML and work with ecosystems including SBML Level 3 Packages, FAIR data principles, Minimum Information About a Simulation Experiment, and repositories like BioModels Database and JWS Online. Key conceptual building blocks connect to experimental resources such as Gene Expression Omnibus, Protein Data Bank, ChEMBL, KEGG, and Reactome to enable annotations and cross-references that improve reuse by projects at European Bioinformatics Institute, NCBI, EMBL-EBI, and ELIXIR.

Levels and Versions

SBML is organized into Levels and Versions that introduce capabilities and optional extensions known as packages. Level 1 and Level 2 established core reaction network semantics used by early tools like Jarnac and BioSPICE; Level 3 introduced modular packages enabling features such as layout, flux balance constraints, hierarchical model composition, and spatial representation used in tools like COPASI, CellDesigner, Virtual Cell, and Morpheus. Standards efforts involved stakeholders from FAIRsharing, COMBINE, COMBINE Archive, Systems Biology Ontology, and funding bodies including European Commission and National Science Foundation.

Software and Tools

A wide range of software supports SBML import/export and simulation: numerical tools such as COPASI, Tellurium, MATLAB SimBiology, Julia DifferentialEquations.jl ecosystems, and visualization tools like CellDesigner and Cytoscape. Libraries and bindings include LibSBML (C/C++), sbmljs, python-libsbml, and interfaces for R Project, Perl, Java, Julia, and Python that enable integration with platforms like Galaxy, Jupyter Notebook, Docker, Singularity, and cloud providers like Amazon Web Services and Google Cloud Platform.

Applications and Use Cases

SBML is used to share kinetic models of signaling pathways studied at Broad Institute, Salk Institute, Cold Spring Harbor Laboratory, and Max Delbrück Center, metabolic network reconstructions from UC San Diego and ETH Zurich, pharmacokinetic/pharmacodynamic models in collaborations with Food and Drug Administration and European Medicines Agency, and multi-scale modeling projects tied to Physiome Project and Virtual Physiological Human. Domains include enzymology studies referencing PDB, systems pharmacology in industry partners like Pfizer and Novartis, and synthetic biology workflows from MIT Synthetic Biology Center and Harvard Wyss Institute.

Standards and Community Governance

Community steering involves organizations and initiatives such as COMBINE, FAIRsharing, ELIXIR, European Molecular Biology Laboratory, National Institutes of Health, and Wellcome Trust. Specification development has seen contributions from research groups at University of Cambridge, ETH Zurich, Yale University, University of Tokyo, and software developers behind LibSBML and SBML ODE solvers. Outreach, workshops, and standards harmonization occur alongside related efforts like Systems Biology Ontology, Minimum Information About a Simulation Experiment, BioPAX, SBGN, and repositories including BioModels Database and JWS Online.

Category:Bioinformatics