search

obophenotype / cell-ontology

An ontology of cell types
https://obophenotype.github.io/cell-ontology/
Creative Commons Attribution 4.0 International
145 stars 49 forks source link

Proposal: obsolete meso/metanephros subtypes for many kidney cell types #512

Open cmungall opened 6 years ago

cmungall commented 6 years ago

From KPMP meeting

E.g. http://purl.obolibrary.org/obo/CL_0000653

image

This is generally confusing since

  1. the meso form is likely rudimentary / non functional in mammals
  2. better to post-compose these
cerivs commented 6 years ago

Meso form would be the form found in fish.

cmungall commented 6 years ago

Yep. Do you typically use the pre-composed subclass or annotate to the main cell type? This Q applies to other cell types too. Quick browse of ZP shows a mix but I haven't rigorously checked.

The danger is that some proportion of curators will annotate to cell type C since it's "obvious" what the kidney stage is, in other cases a ZFIN/MGI curator will annotate to "meso-C" or "meta-C", since they have been trained to be as specific as possible, and we end up with a bit of an ad-hoc mix of specificity. Of course we can infer the subtyping if we also have stage info (usually) but this complicates the overall pipeline

cerivs commented 6 years ago

We should be making sure our curators have and use the specific term type. I would prefer our annotations go to the most specific cell type and have automated processes pull the annotations together rather than requiring users to realize that our annotations are to cell types in different anatomical structures (pronephros, mesonephros and metanephros). I think we need the general cell type for the cross species or cross stages comparison but we might need the specific terms so inherent differences based on developmental stage or species are simpler to determine.