Import of ENZYME data into the ConceptWiki and its representation as RDF

Solutions to the classic problems of dealing with heterogeneous data and making entire collections interoperable while ensuring that any annotation, which includes the recognition-and-reward system of scientific publishing, need to fit into a seamless beginning to end to attract large numbers of end users. The latest trend in Web applications encourages highly interactive Web sites with rich user interfaces featuring content integrated from various sources around the Web. The obvious potential of RDF, SPARQL, and OWL to provide flexible data modeling, easier data integration, and networked data access may be the answer to the classic problems. Using Semantic Web technologies we have created a Web application, the ConceptWiki, as an end-to-end solution for creating browserbased readwrite triples using RDF, which focus on data integration and ease of use for the end user. Here we will demonstrate the integration of a biological data source, the ENZYME database, into the ConceptWiki and it’s representation in RDF.

💡 Research Summary

**
The paper presents an end‑to‑end solution for integrating heterogeneous biological data into a web‑based semantic platform called ConceptWiki, and for exposing the integrated data as RDF triples. ConceptWiki is described as an open‑access, wiki‑style repository that stores concepts, their synonyms in multiple languages, and basic annotations such as definitions. Each concept is identified by a Universally Unique Identifier (UUID) that is deliberately opaque – it carries no intrinsic meaning and never changes even if the underlying information is updated. This design choice eliminates identifier clashes, simplifies version control, and enables seamless linking with other resources that also use UUIDs.

The authors illustrate the import workflow using the ENZYME database, a curated collection of enzyme information indexed by Enzyme Commission (EC) numbers. The original ENZYME flat file is first transformed into XML. An import script parses the XML, queries the ConceptWiki backend for concepts that already contain the same EC number as a synonym, and then compares the stored ConceptWiki data with the ENZYME record. If the ENZYME entry does not exist in ConceptWiki, a new concept is created; if it exists but differs, the stored record is updated. The user interface reflects these changes through an “authority checkbox” that indicates whether a piece of information is still supported by ENZYME or has been superseded.

In the RDF representation, each triple follows the classic subject‑predicate‑object pattern, but both subject and object are expressed as ConceptWiki UUID URIs, while the predicate is a dedicated URI that explicitly describes the relationship (e.g., hasSynonym, hasFunction). Unlike ordinary HTML hyperlinks, RDF predicates are semantically labeled, making the relationship machine‑readable and suitable for SPARQL queries or OWL reasoning. The paper shows an example where the enzyme “Aldehyde reductase” (EC 1.1.1.1) is linked to the function “sorbitol biosynthetic process” through a newly created triple that combines imported ENZYME data with user‑generated connections.

A key feature of ConceptWiki is its user‑centric editing environment. Through simple drop‑down menus, scientists can create new triples by linking any two concepts. Every newly created triple is automatically attributed to the contributing scientist, displaying their name and listing the triple on a personal page. This attribution mechanism is intended to provide scholarly credit for data curation activities, addressing the broader issue of recognizing contributions that fall outside traditional publication models.

The authors argue that the combination of opaque UUIDs, RDF‑based modeling, and an intuitive web interface makes ConceptWiki a powerful platform for building “semantic mash‑ups.” It enables the integration of diverse terminologies (UMLS, SwissProt, Medline, and future ChemSpider data) into a single, queryable knowledge graph. The RDF output is fully linked to related concepts, turning the wiki into a machine‑processable semantic network. The paper concludes that ConceptWiki lowers the barrier for creating richer web‑based knowledge resources, promotes equitable recognition of contributors—especially those from under‑represented groups—and sets the stage for more advanced semantic applications such as ontology‑driven reasoning and large‑scale data mining. Future work includes performance optimization for bulk imports and extending the OWL layer to support more expressive logical constraints.