The Pistoia Alliance was pleased to collaborate with the Innovative Medicines Initiative (IMI), Virtual Physiological Human (VPH) Network of Excellence, and Computational Modeling in Biology Network (COMBINE) communities to hold a biomedical data and model interoperability workshop on 28-29 March 2011.
Mathematical modeling communities in the domains of systems biology, bioengineering, and pharmacometrics represent an active and diverse network of biomedical scientists that studies human biology by computational means. This diversity is the basis for an impressive breadth of modeling and simulation approaches that generate large volumes of potentially shareable data and model resources (DMRs). Although, in principle, such resources are re-usable by the discovery, development and healthcare communities, in practice few can currently be shared due to a number of obstacles:
- Concerns about clinical data confidentiality
- Limitations due to model intellectual property
- The lack of consistent annotation protocols for biological concepts that are implicitly represented in both clinical and pre-clinical DMRs
- The absence of open and coherent technical infrastructures that allow community users to apply annotation in a consistent manner using standard terminologies/ontologies and subsequently query and reason logically over these distributed annotated repositories
The VPH’s RICORDO (started Feb 2010) and the IMI’s Disease/Drug Modeling Resource (commencing early in 2011) aim to develop protocols and computational infrastructures to support semantic interoperability of biomedical DMRs. The two projects have a number of operational goals in common, specifically:
- Building a communal standard set of ontologies for representing biological structure across different scales (i.e. small molecules to gross anatomy); biological qualities observed in the lab or clinic (e.g. pressure, mass, concentration, etc.); and units of measurement
- Assisting and educating the community in targeted ontology development and semantic interoperability
- Developing tools that make use of the above ontologies to annotate VPH and IMI data and models applying automated ontology-based reasoning methods on DMR annotations to identify novel relationships between existing resources
- Deploying infrastructure in support of sharing and querying annotations, as well as providing the basis for automated analysis and modeling workflows
At the workshop in March, participants discussed how to identify practical causes of and solutions to semantic interoperability obstacles. In particular, are aim is to work with modelers, knowledge managers, and data stewards from the EFPIA partners in the Disease/Drug Modeling Resource (i.e. AstraZeneca, GlaxoSmithKline, Lilly, Merck, Novartis, NovoNordisk, Pfizer, Hoffmann-LaRoche, Servier, and UCB) to establish the communal requirements and metadata standards for the efficient application, sharing, and querying of DMR annotations. These groups presented and demonstrated interoperability toolkits in current use within both industry and academic communities.
This meeting was sponsored by the European Bioinformatics Institute Industry Programme and was held at the Wellcome Trust Genome Campus in the UK.