Several potential modeling issues in Jan 2016 UBERON Release

[ChrisOchs]

As part of an on-going research project my lab is currently reviewing several ontologies (e.g., SNOMED, NCIt, GO, others) for modeling issues. A domain expert working with us performed a preliminary review of a small sample of UBERON classes taken from the Jan 2016 release. We were asked to post the identified issues on the tracking to obtain feedback.

The issues are listed below in-order according to the class name. I did not personally identify these issues but I can contact our domain expert to get more information or to clarify.

1st arch mesenchyme from neural crest

remove restriction develops from some neural crest

The range of this restriction is more general than the other develops from restriction, develops from some rhombencephalon neural crest

3rd arch mesenchyme from neural crest

Same as above. Additionally, modeling is not consistent with 1st arch mesenchyme from neural crest, as this class has a has part restriction to mesenchyme from rhombencephalic neural crest but the other class does not.

anatomical lobe Subclasses appear to be incomplete. For example, Lobe of liver is not a subclass (more specifically, these classes are missing anatomical lobe as a superclass.

anatomical wall Subclasses also appear to be incomplete, depending on definition. Sublasses such as conjuctiva and the tympanic membrane are questionable.

aponeurosis Similarly to the above two classes, this class appears to be missing subclasses (e.g., abdominal, lumbar, plantar)

basioccipital bone

Remove restriction: Part_of Chondrocranium does not seem correct for a bone. It develops_from the condrocranium but the bone is not part_of it. (The domain expert was not sure if this is true in non-human species)

Change restriction: Part_of Occipital region should be changed to Part_of Occipital bone which is part of the occipital region

[uberon]

These are great, we will attend to these next week.

[cmungall]

1st arch mesenchyme from neural crest

remove restriction develops from some neural crest

The range of this restriction is more general than the other develops from restriction, develops from some rhombencephalon neural crest

This and similar issues should have been fixed by #1219

Detailed explanation: the equivalence axiom uses the more general class (which makes it a stronger axiom), the subclass axiom (a hidden GCI) is more specific. In previous releases all equivalence axioms were relaxed to SubClassOf axioms. These relaxed axioms are now removed if they are redundant.

But note the equivalence axiom is retained. Depending on how your browser renders uberon, there may be apparent redundancy, but there will be no actual redundancy.

3rd arch mesenchyme from neural crest

Same as above. Additionally, modeling is not consistent with 1st arch mesenchyme from neural crest, as this class has a has part restriction to mesenchyme from rhombencephalic neural crest but the other class does not.

The part_of has been removed

anatomical lobe Subclasses appear to be incomplete. For example, Lobe of liver is not a subclass (more specifically, these classes are missing anatomical lobe as a superclass.

Fixed

Also added this as parent for UBERON:0000101 ! lobe of lung

In future 'anatomical lobe' may be removed from uberon

anatomical wall Subclasses also appear to be incomplete, depending on definition. Sublasses such as conjuctiva and the tympanic membrane are questionable.

This abstraction comes from FMA. It is consistent with the FMA definition:

def: "Organ component adjacent to an organ cavity and which consists of a maximal aggregate of organ component layers." [FMA:82482]

However, it defies intuition. We will look into a more anatomist-friendly definition of anatomical wall

aponeurosis Similarly to the above two classes, this class appears to be missing subclasses (e.g., abdominal, lumbar, plantar)

We don't have these classes, so is this more of a new term request?

The following are classified as aponeuroses

UBERON:0006661 ! epicranial aponeurosis UBERON:0014780 ! palatine aponeurosis

basioccipital bone

Remove restriction: Part_of Chondrocranium does not seem correct for a bone. It develops_from the condrocranium but the bone is not part_of it. (The domain expert was not sure if this is true in non-human species)

"Chondrocranium" is a problematic label. We use it consistent with the evo-devo literature to mean "that part of the neurocranium formed by endochondral ossification and comprising the bones of the base of the skull"

We have the class "cartilaginous chondrocranium" to refer to what is more typically called "chondrocranium" in the human literature

We'll look at making the nomenclature less confusing

Change restriction: Part_of Occipital region should be changed to Part_of Occipital bone which is part of the occipital region

Note this is not necessarily true outside mammals.

We do have a Taxon-specific GCI:

('basioccipital bone' and part-of some 'mammalia') SubClassOf part-of some 'Occipital bone' ## ISBN10:0073040584

This may not be visible in all environments. I recommend Protege 5 here. See:

https://github.com/obophenotype/uberon/wiki/Evolutionary-variability-GCIs

[gouttegd]

WARNING: This issue has been automatically closed because it has not been updated in more than 3 years. Please re-open it if you still need this to be addressed addressed addressed – we are now getting some resources to deal with such issues.