Welcome and Introduction to CellML

CellML is designed to support the definition and sharing of the models of biological process. It is intended to provide the consistency in mathematical representation. CellML encourages model evolution and reuse. It also provides a representation in a form (XML) that is both human and computer readable. CellML project started in 1999, around the same time when SBML commenced. CellML and SBML have different emphasis; SBML is designed to represent biochemical reaction while purpose of CellML is to provide a generic framework to represent mathematical models of biological and physiological systems. CellML includes information about model structure, mathematics and metadata. CellML includes mathematics and metadata by leveraging existing XML based standards such as MathML. CellML philosophy is revolving around components and model structure is based upon connected components. CellML components provide abstraction and encapsulation using public and private interfaces. A model is root element of a CellML document. Components are connected using model mathematics defined using content MathML. CellML import element allows model reuse and new models can be created by combining existing models into model hierarchy. Currently CellML model repository have more than 350 published model curated manually, covering different aspects of biological systems such as cell signalling, electrophysiology, metabolic networks, calcium dynamics, endocrine systems, cell cycles, smooth and skeletal muscles. Model curation is ongoing process and models are tagged with an attribute which indicates level of curation (Level 0, 1, 2, 3). In future CellML models can be decomposed into reusable functional unites, resulting in a library of decomposed units which can further be used to build new models by pulling functional units from library by implementing CellML imports. CellML is useful to exchange and implement time independent descriptions of models, but higher level knowledge is not captured in the process of constructing the CellML representation. Biological information may be included informally as metadata, but this approach is insufficiently rigorous. CellML need a linking mechanism to associate the CellML entities to well defined representation of knowledge such as ontology. Such a mechanism can exploit existing ontologies at various levels such anatomical (GO, FMA), gene regulation and pathway (GO, GONG, BioPAX), gene expression (GO, GANG) aside from new ontologies developed as part of physiome project. There are variety of tools such as COR, PCEnv, CMISS, insilicoIDE, Virtual Cell, JSim those can be used for authoring CellML models and further simulating them. A new tool OpenCell is evolving which integrates best features from COR and PCEnv. Currently several changes are proposed for next version of CellML specification such as removal of reaction element, introducing concept of secondary specifications.