drains/supplies - should we represent at both the organ level and the organ part level? Should we use inference to do this?

@emquardokus wrote:

are we looking for a relationship to the organ per se or to the hilum itself, which is a part of the organ or are you lumping it altogether? The relationship I see described in the literature in anatomical terms is all the vessels entering or exiting via hilum. On the other hand the contents (blood or lymph) of said vessels can be draining or supplying said organ. Maybe it's a biology thing as I see this as 2 separate relationships... physical vessel to organ vs functional thing what it's doing.

On the general issue:

• I'd like to document what the CCF use cases are for representing connection/supply/drains relationships in CCF.owl and wether this would require relationships both at the organ level and at the organ part level.
• Rather than adding two relationships to Uberon for drains/supplies covering organ and organ part, I'd prefer to use reasoning. A simple property chain can be used to say that if X drains/supplies Y and Y part_of Z then X drain/supplies Z. That might be too strong though as part_of is transitive (every vessel supplies the whole body...). We could use GCIs to restrict this to the organ level instead. (drains some (part_of some kidney) subClassOf drains some kidney).