Phenoscape

Phenoscape models logical homology relationships to integrate comparative anatomical phenotypes and genetic data for evolutionary inference.

Key Features:

• Logical homology modeling: Evaluates logical homology using the Reciprocal Existential Axioms (REA) model and the Ancestral Value Axioms (AVA) homology model.
• Homology types represented: Represents historical homology and serial homology across anatomical structures.
• Phenoscape Knowledgebase integration: Uses the Phenoscape Knowledgebase, compiling homology assertions from literature alongside phenotypic data for skeletal elements of vertebrate fins and limbs.
• AVA model behavior: The AVA model returns user-expected results as well as subclasses of search terms and any superclasses where a homology relationship has been asserted.
• REA model performance: The REA model satisfies five of seven competency queries, indicating limitations in comprehensive homology queries due to OWL reasoning constraints.
• Ontology-based integration: Integrates homology reasoning into ontology-based analyses to support synthetic supermatrix construction and candidate gene discovery.

Scientific Applications:

• Comparative anatomy and evolutionary inference: Integrate and compare evolutionary findings across diverse studies and organisms and relate anatomical changes to genetic underpinnings.
• Synthetic supermatrix construction: Support building synthetic supermatrices for phylogenetic and comparative analyses.
• Candidate gene discovery: Prioritize candidate genes linked to anatomical evolution by linking homology assertions with genotype–phenotype data.
• Ontology reasoning assessment: Test and evaluate OWL-based reasoning on homology queries using competency queries and model comparisons (REA vs. AVA).

Methodology:

Compiles homology assertions from literature and phenotypic data for skeletal elements of vertebrate fins and limbs in the Phenoscape Knowledgebase; evaluates the Reciprocal Existential Axioms (REA) and Ancestral Value Axioms (AVA) models using OWL reasoning, with the AVA model returning subclasses and superclasses of asserted homologies and the REA model addressing five of seven competency queries.