Implementation of Zebrafish Ontologies for Toxicology Screening

Area of Science:

• Toxicology
• Zebrafish research
• Data science

Background:

• Toxicological evaluations using zebrafish (Danio rerio) report altered phenotypes.
• Variability in reported phenotypes and inconsistent data annotation necessitate harmonization.
• Data science approaches can reconcile differences in measurements across studies.

Purpose of the Study:

• To evaluate the impact of standardized phenotype terminology on data consistency in zebrafish toxicology.
• To assess the effectiveness of the Zebrafish Phenotype Ontology in improving data annotation.

Main Methods:

• Zebrafish researchers annotated larval images for morphological malformations in two surveys.
• Survey 1: Annotation using individual researchers' terminology.
• Survey 2: Annotation using terms from the Zebrafish Phenotype Ontology.

Main Results:

• Use of ontology terms increased consistency in phenotype reporting.
• Ontology-based annotation decreased ambiguity compared to individual terminology.
• Preliminary results suggest improved agreement and repeatability.

Conclusions:

• Standardized terminology and common data standards reduce heterogeneity in reported terms.
• Ontology application enhances agreement and repeatability between laboratories.
• A zebrafish phenotype atlas is proposed to foster interoperable and computable data.