Unsats upon merging FBbt with Uberon

[gouttegd]

Merging FBbt with the taxon-neutral ontology Uberon still results in ~100 unsatisfiable classes. They are all caused by the fact that 7 FBbt classes are simultaneously part of both the central nervous system and the peripheral nervous system (something that Uberon does not allow, hence the unsatisfiability). Those classes are:

• adult stomodeal nerve,
• dorsal margin photoreceptor cell R7,
• dorsal margin photoreceptor cell R8,
• embryonic/larval frontal nerve,
• embryonic/larval recurrent nerve,
• larval esophageal nerve 1,
• larval esophageal nerve 2.

adult stomodeal nerve

Explicitly classified as both nerve (which in FBbt, and in Uberon, is defined as being part of the PNS) and part of some adult hypocerebral ganglion (which is itself classified as part of the adult CNS).

The classification as a part of the hypocerebral ganglion seems wrong judging by the 1994 description from Rajashekhar and Singh (FBrf0080339), since the nerve is not entirely contained within the HCG. The relation should probably be innervates some 'adult corpus cardiacum' instead.

dorsal margin photoreceptor R7

(The problem with dorsal margin photoreceptor cell R8 is exactly the same.)

Classified as part of some dorsal margin ommatidium, which is ultimately (through both ommatidium and dorsal rim area) part of some retina, which is itself part of the adult PNS.

And also classified as part of some dorsal rim area optic column, which is ultimately a synaptic neuropil, which is part of some neuropil, which is itself part of the CNS.

The latter classification is probably wrong, because:

• as a synaptic neuropil, the dorsal rim area optic column is expected to “consist largely of synapsing neuron projections”, i.e. be devoid of neuronal bodies, so it seems unlikely that photoreceptor cells would be entirely contained within it;
• the parent term dorsal margin photoreceptor (under which 'dorsal margin photoreceptor cell R7' is automatically classified) is said to be 'sending synaptic output to region' some 'dorsal rim area optic column', which is more consistent with the idea of a receptor cell that is located in the retina and that projects to an optic column.

Suggested fix here is to simply remove the 'part of' some 'dorsal rim area optic column' classification. (No need to replace it with 'sends synaptic output to region' some 'dorsal rim area optic column', since that relation is inherited from 'dorsal margin photoreceptor'.)

embryonic/larval frontal nerve

(The problem with 'embryonic/larval recurrent nerve', 'larval esophageal nerve 1', and 'larval esophageal nerve 2' is the same.)

Ultimately classified as a nerve and therefore as part of the PNS.

And also directly classified as part of some embryonic/larval stomatogastric nervous system (SNS), which is part of the embryonic/larval CNS.

To solve this particular contradiction, I’m inclined to drop the relationship with the stomatogastric nervous system. All those nerves are already said to be innervating one or several of the ganglions that make up the embryonic/larval SNS (which are themselves part of that system), so they would still be (indirectly) linked to the SNS without being “part” of it.

[gouttegd]

Following Uberon’s last release there are also 845 new unsats that are caused by the fact that insect dorsal ectoderm derivative in Uberon is classified as a multi-cellular anatomical structure, whereas our our dorsal ectoderm derivative is classified only as an anatomical entity.

I think our classification is correct (a “dorsal ectoderm derivative” can be a single cell, such as a histoblast, or even an acellular structure, such as an apodeme), so that is something that will need fixing on the Uberon side.