Challenges in developing multi-species anatomy ontologies

Representing taxonomically variable part_of relationships for an entity

See related TAO tracker request

An entity may have taxonomically variable part_of relationships. For example, vertebra 1 may or may not be part_of the Weberian apparatus. (The Weberian apparatus is an anatomical cluster consisting of the first four vertebrae (additional vertebra in some taxa) and associated structures which connect the gas bladder to the inner ear. It is diagnostic of the Otophysi (Cypriniformes, Characiformes, Gymnotiformes, Siluriformes). Some species have vertebra 5 and/or 6 as part of Weberian apparatus). How to represent taxonomically variable part_of information in a multi-species ontology?

1. Multiple vertebra 1 terms as children of “vertebra”

precaudal vertebra
     <--is_a vertebra 1
     <--is_a vertebra 1 part of Weberian apparatus

The term “vertebra 1 part of Weberian apparatus” has relationship part_of: Weberian apparatus

- Requires curator to choose the taxonomically correct term at time of annotation

- Search on “vertebra 1” will bring back results for all ostariophysans only if homologous_to statement is used:

"vertebra 1" in Anotophysi homologous_to "vertebra 1 part of Weberian apparatus in Otophysi

2. Vertebra 1 term with child “vertebra 1 part_of Weberian apparatus”

precaudal vertebra
     <-- vertebra 1
           <-- vertebra 1 part_of Weberian apparatus

- Also requires curator to choose the taxonomically correct term at time of annotation

- Search on "vertebra 1 will bring back results for all ostariophysans, regardless of whether homologous_to statement is made

3. part_of variability as a phenotype statement

E: vertebra; Q: ? part_of; Related E: Weberian apparatus; Taxon: Otophysi

- Problem: this is not a species-level annotation - What is appropriate quality?

Representing serial homologues

neural arch 2 and intercalarium are children of neural arch