integumental system

[tfhayamizu]

The integumental system (UBERON:0002416) should NOT be a mesoderm-derived structure (UBERON:0004120; "implied link automatically realized") because (at least in most, if not all, cases) most of the integument is thought to be derived from ectoderm; i.e. does NOT develop from mesoderm (UBERON:0000926).

[tfhayamizu]

I also question whether it should be considered to be disjoint from the "entire sense organ system" and "immune system".

[cmungall]

You are right, the link should not be there. It has been removed (I will close this ticket when the fix is in the release version).

The class in question here is integumental system which is broader than integument or skin and seems to include a broad mix of ectodermal and mesodermal derivatives. The best thing to do here is obviously to make more concrete statements at the level of individual tissues and layers. E.g epidermis-ectoderm, dermis and hypodermis-mesoderm.

Regarding disjoint_from - not that this means class disjointness, so they are in fact disjoint. The semantics of this are defined by the owl specification, and it means that nothing is both an integumental system and an immune system. It does not mean that nothing can be part of both. We do have some statements of the latter form in Uberon. The way we write this in OWL is:

(part_of some brain) DisjointWith (part_of some spinal cord)

We have a shortcut relation "spatially_disjoint_from" in the uberon obo file.

An interesting unresolved question is how much of the tissue has to have its origin the the precursor for a develops_from link to be justified (and whether this is mass, # of cells, ...). Ideally we would have as many assertions as possible at the level of cell types in individual species, and we would dynamically infer the tissue-level relationships in a multi-species context from here?

[cmungall]

fixed in release version. Thanks again.