A framework for annotating and visualizing cellML models

Abstract

The Physiome Project was established to develop tools for international collaboration and sharing physiological knowledge in the form of biological models and experimental data. The CellML language was developed in response to the need for a high-level language to represent and exchange mathematical models of biological processes. The language provides a flexible framework for describing the dynamics of biological processes but does not explicitly lend itself to capturing the underlying biological concepts such as the entities and processes that these models represent. The relationship between the biological process and the mathematical model describing the biological process is also often complex. This makes it difficult to see the biological concepts which the CellML structures represent. A framework which supports visualizing the biological concepts and its relationship to the underlying CellML model would provide a very useful toolset for understanding the biological concepts modeled in CellML. The CellML models need to be annotated with biological concepts in order to provide the machine interpretable data for generating a visual representation. We have developed an ontological framework which can be used to explicitly annotate CellML models with physical and biological concepts, a method to derive a simplified biological view from the annotations, a visual language for representing all biophysical processes captured in the CellML models, and a method to map the visual language to the ontological framework in order to automate the generation of visual representations of a model. The proposed method of model visualization produces a result that is dependent on the structure of the CellML models which requires modelers to structure the model in a way that best describes the biophysical concepts and abstractions they wish to demonstrate. Our argument is that this leads to a best practice approach to building and organizing models. As a part of this research, a software tool for visualizing CellML models was developed. This tool combines the visual language and the ontologies to generate visualizations that depict the physical and biological concepts captured in CellML models and enables different communities in diverse disciplines to more easily understand CellML models within the biological domain they represent. As research continues, with further improvement to the framework it would be possible to visually construct composite CellML models by selecting high level biological concepts.