[Obsolete] 'neural cell'

[dosumis]

CL term

neural cell http://purl.obolibrary.org/obo/CL_0002319 A cell that is part of the nervous system.

EquivalentTo: (animal cellandpart ofROsomenervous systemUBERON)

Suggested replacement

This is a rare case where I think no replacement can be provided

Reason for deprecation

This is so general & trivial that it is useless. In work with PCL it is leading to distracting automated classifications that are interfering with expert review.

[rays22]

I think it causes problems if there is no replacement for CL:0002319. CL:0002319 is used by 30 different ontologies, for example, as a component term to define classes.

[ghost]

@dosumis, given the concern noted in the comment above, would you still like to proceed with this obsoletion?

[dosumis]

I think it causes problems if there is no replacement for CL:0002319. CL:0002319 is used by 30 different ontologies, for example, as a component term to define classes.

I think we need to work out if it is being used directly by any ontologies, or whether it just turns up in import modules.

[dosumis]

The only way I know to do this is right now is with OntoBee (We should ask for this functionality with the OLS API).

The majority of use is in imports, but looking through I can see at least a couple of uses:

https://ontobee.org/ontology/ECO?iri=http://purl.obolibrary.org/obo/CL_0002319

3D neural cell structure determination assay equivalentClass : 3D cell structure determination assay and ( has_specified_output some 3D structural organization datum) and ( has_specified_output some is about some neural cell)

https://ontobee.org/ontology/UBERON?iri=http://purl.obolibrary.org/obo/CL_0002319

brain subClassOf : composed primarily of some neural cell nervous system subClassOf : composed primarily of some neural cell

Need more complete exploration for other direct uses.

We can choose to drop from Uberon. We could recommend that ECO use "cell and 'part of' some brain" in place of neural cell...

[balhoff]

Here are a few ontologies with direct uses in existential restrictions: https://api.triplydb.com/s/89q9DYLEQ

• obo:hp/hp-base.owl
• obo:oba/oba-base.owl
• obo:obi/obi-base.owl
• obo:uberon/uberon-base.owl

[jahilton]

'neural cell' seems to be a general term and thus, can be useful in grouping its descendants. I believe that removing this term would set a precedent throughout CL and require a very substantial overhaul. Just a few examples: 'kidney cell' [CL:1000497], 'motile cell' [CL:0000219], 'connective tissue cell' [CL:0002320] Or is there something specific to 'neural cell' that I am missing relative to those others?

[ANiknejad]

adding my 2-cents thoughts here: maybe the question is if broadly defined cells can be used for single-cell transcriptomics cell type identification, when a cell type can not be defined but still we know coming from a specific tissue (brain, kidney, connective tissue). Can we safely report 'connective tissue cell' [CL:0002320] or 'kidney cell' [CL:1000497] or 'neural cell' [CL:0002319] for such unknown cell types ? are T cells and any other invading cells considered as part of this tissue too ? CellXGene likes reporting such unknown cells as 'native cell' but I see that this class is also under discussion for obsoleting!

https://github.com/chanzuckerberg/single-cell-curation/blob/main/schema/4.0.0/schema.md#integration-metadata

[ghost]

CellXGene likes reporting such unknown cells as 'native cell' but I see that this class is also under discussion for obsoleting!

To clarify, 'native cell' has been obsoleted as of the latest CL release.

[ANiknejad]

yes, thank you Bradley, I saw also the issue 622 in the chanzuckerberg github, talking about the 2 options to replace ‘native cell’

[github-actions[bot]]

This issue has not seen any activity in the past 6 months; it will be closed automatically in one year from now if no action is taken.