Increasing NCBO BioPortal and CEDAR Synergy for BD2K

The NCBO BioPortal provides Web services for over 500 biomedical ontologies, allowing investigators to annotate and retrieve data, generate value sets, and perform advanced analytics of a wide range of biomedical and clinical data. BioPortal provides core services for many CEDAR metadata activities, and also heavily serves the wider BD2K community—seven out of eight centers responding to the BD2K metadata survey cited BioPortal as an essential resource for their work. 

We have completed some long-planned features enabling or optimizing CEDAR’s use of BioPortal. With this year’s supplement, we have begun pursuing enhancements to maximize BioPortal’s value to CEDAR and advance BioPortal’s integration with other BD2K programs and services. We present the work accomplished, and enhancements that are being made available over the course of this supplement.

The first set of completed features advanced CEDAR‘s usage of BioPortal concepts and value sets. We made subtle improvements to BioPortal’s ontology presentation services and added features for accessing and extending value set services.

We also began enhancing BioPortal’s term-specific services for use by CEDAR and others. We have prototyped a concept-centric view of ontology data, providing for a given term all significant information available across all BioPortal ontologies as well as other relevant sources. Using this baseline, we plan to provide a “best-term” identification service, taking into account more contextual information specific to term discovery; and a query-expansion service. 

As CEDAR adds property relations for its fields and terms, we will advance BioPortal’s handling of ontology properties (the formal relationships between terms), extending the BioPortal API to better handle and retrieve ontology properties. We’ll also be increasing BioPortal’s metadata-handling capabilities.

Finally, we intend to harden BioPortal’s capacity for complex queries by CEDAR, both by optimizing query handling (partially completed), and developing application-level test suites around CEDAR’s most important queries.