Annotation of SBML Models Through Rule-Based Semantic Integration

Abstract

Abstract*Motivation:* The creation of accurate quantitative Systems Biology Markup Language (SBML) models is a time-intensive, manual process often complicated by the many data sources and formats required to annotate even a small and well-scoped model. Ideally, the retrieval and integration of biological knowledge for model annotation should be performed quickly, precisely, and with a minimum of manual effort. Here, we present a method using off-the-shelf semantic web technology which enables this process: the heterogeneous data sources are first syntactically converted into ontologies; these are then aligned to a small domain ontology by applying a rule base. Integrating resources in this way can accommodate multiple formats with different semantics; it provides richly modelled biological knowledge suitable for annotation of SBML models. *Results:* We demonstrate proof-of-principle for this rule-based mediation with two use cases for SBML model annotation. This was implemented with existing tools, decreasing development time and increasing reusability. This initial work establishes the feasibility of this approach as part of an automated SBML model annotation system.*Availability:* Detailed information including download and mapping of the ontologies as well as integration results is available from "http://www.cisban.ac.uk/RBM":http://www.cisban.ac.uk/RBM