A Systematic Analysis of Term Reuse and Term Overlap across Biomedical Ontologies

Area of Science:

• Biomedical Informatics
• Ontology Engineering
• Computational Biology

Background:

• Ontology reuse is encouraged for semantic interoperability and cost reduction.
• Biomedical ontologies are crucial for organizing complex biological data.

Purpose of the Study:

• To assess the extent of term reuse and overlap in biomedical ontologies.
• To identify patterns and challenges in current ontology term reuse practices.

Main Methods:

• Analysis of a corpus of biomedical ontologies from the BioPortal repository.
• Examination of reuse and overlap constructs, including clustering and Protégé plugin logs.
• Development of a web application for visualizing reuse and suggesting terms.

Main Results:

• Approximate term overlap is 25-31%, but actual term reuse is less than 9%.
• Reused terms exhibit high semantic similarity (>90%), often as sibling or parent-child nodes.
• Identified error patterns suggest unintended reuse due to representation differences.

Conclusions:

• Ontology development methodologies should prioritize enhancing term reuse.
• Semi-automated tools offering personalized recommendations are needed to support ontology engineers.
• Improving term reuse can significantly boost semantic interoperability and reduce development costs.