Transitive homology

From phenoscape

This page explores the transitive closure of the homologous_to relation. Currently in OBD this is a binary class-level relation that holds between two taxon-specific anatomical entities.

I propose the following expansion to OWL-DL instance-level quantified axioms:

 homologous_to(A,B) ==>
 Class: A
 SubClassOf: descendedFrom some (hasDescendant some B)
 Class: B
 SubClassOf: descendedFrom some (hasDescendant some A)

If we make the simplifying assumption that directlyDescendsFrom is functional then it follows that class-level homology is transitive. What are the consequences of making this strong assumption? We can explore this with existing phenoscape data.

High Level Overview

This is a high level part_of view of all homology relationships, projected onto TAO:

Error creating thumbnail: Unable to save thumbnail to destination
  • homology statements are generalized to the TAO, taxa not shown
  • clusters of homology relationships are shown in 4 distinct boxes
  • vertebra clustered into a box
  • only relations shown are homology and part_of

Here we can see there are 4 distinct clusters, according to the vertebra number affected

Otophysi intercalarium

this is annotated to be homologous to the neural arch 2 and vertebra 2 in Teleosti. From this, what can be said of the relationship between these two entities?

Error creating thumbnail: Unable to save thumbnail to destination

Note that for compactness reasons this diagram doesn't show the part_of relations.

  • intercalarium part_of v2

presumably there should be some kind of relationship between "rib of v2" and "v2"

Otophysi tripus

this is annotated to be homologous to (i) the fusion of parapophysis + rib of v3 and (ii) the fusion of parapophysis + rib of v3 + rib of v4 in Teleosti. What can be said from this of the relation between (i) and (ii)

Error creating thumbnail: Unable to save thumbnail to destination