Ontology Development in Biosciences

Abstract

Biomedical ontologies provide standardised vocabularies and logical frameworks for representing biological knowledge, enabling data integration, automated reasoning, and cross-database interoperability, yet ontology development practices vary substantially across bioscience domains and communities. We evaluated 216 ontology development programmes across centres in Estonia and Austria between 2018 and 2022, spanning five development approach categories: consensus-driven expert curation, automated term extraction from literature, AI-assisted ontology alignment, community-contributed collaborative editing, and reference ontology harmonisation frameworks. A Bioscience Ontology Development Index (BODI) was constructed from five sub-scores -- term coverage and granularity, logical consistency and reasoning support, community adoption breadth, update and maintenance responsiveness, and interoperability with existing ontologies -- with weights from regression against sustained adoption. BODI correlated with adoption at r = +0.84 and discriminated adopted from non-adopted ontologies with an AUC of 0.884. Consensus-driven expert curation scored highest (mean BODI 0.826), while AI-assisted alignment trailed at 0.600. Only 35.6 percent exceeded the 0.75 threshold. Term coverage carried the largest weight (beta = +0.278), followed by interoperability (beta = +0.230).