neuroblast and neural progenitor cell classification in CL

[cmungall]

CL currently has hierarchies for neuroblast (vertebrate) and NPC:

• [] CL:0000055 ! non-terminally differentiated cell "A precursor cell with a limited number of potential fates."
  • [i] CL:0000031 ! neuroblast (sensu Vertebrata) "A cell that will develop into a neuron often after a migration phase."
    • [i] CL:0002362 ! cerebellar granule cell precursor "A cell located in the outermost proliferative zone of the external germinal layer that can differentiate into astroglial cells and granule cells. This cell type is glial fibrillary acidic protein-positive and HNK1-positive."
    • [i] CL:1000042 ! forebrain neuroblast
    • [i] CL:0002676 ! neural crest derived neuroblast "A neuroblast derived from a neural crest cell."
  • [i] CL:0011020 ! neural progenitor cell "An undifferentiated cell that originates from a neural stem cell and has the capacity to generate multiple types of lineage-restricted progenitors but not to self-renew."
    • [i] CL:0000469 ! ganglion mother cell "A neural progenitor cell that is the daughter of a neuroblast (sensu arthopoda). The progeny of ganglion mother cells develop into neurons, glia and (occasionally) epithelial cells."

I'm not sure I could use these definitions to determine if something is a NPC or a NB or both. Suspiciously, the NPC branch only has a single child, which is arthropod-specific.

Our definition of neuroblast differs quite a bit from the more precise wikipedia def from which it was sourced:

In vertebrates, a neuroblast or primitive nerve cell[1] is a postmitotic cell that does not divide further,[2] and which will develop into a neuron after a migration phase. ... Vertebrate neuroblasts differentiate from radial glial cells and are committed to becoming neurons

We also lack the relationship to RGCs.

Note that despite the definition of NPC stating it derives from a neural stem cell, there is no such relationship in CL

Is the concept of neuroblast even useful in vertebrates any more? This recent review of NPC terminology (used in the definition of NPC) barely mentions the term despite talking about NPC cells migrating to the SVZ:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6291443/

I see cellxgene subsets on a lot of these terms - does this mean they are being used to annotate single cell data? are there not more precise concepts that would be more informative?

Some general recommendations

• rather than using a pick and mix of sources for definitions and terminology, we try and focus on recent broad reviews such as PMC6291443
• we have fewer grouping concepts, and less overlapping ways to group things
• all grouping concepts should have a strategy for ensuring consistent and Reasonably Complete population, such as logical definitions, and precise grouping criteria (e.g. "often" is useless in the non-gloss part of the definition)

Specific actions

- [ ] consider placing NB (V) under NPC

• Clarify definition of NPC to make clear it refers to cells that divide to form...
• add GMC develops_from NB (invertebrate)
• consider: add NB develops_from RGC
• add text definition for forebrain NB
• make the definition of NB (V) the same as the wikipedia page it sources (or find a review article)

[dosumis]

Suspiciously, the NPC branch only has a single child, which is arthropod-specific.

It is worth bearing in mind that CL has branches that are hooks for a wider classification of terms in SS ontologies. The CL graph alone is misleading here. There are other cell types that fit this description in invertebrates, e.g. SMCs (Sensory Mother Cells). Not sure whether the mappings are in place for this, but they should be.

[dosumis]

I think an important distinction that is being missed here is that vertebrate NBs are neurons prior to differentiation. The mother cells classified under 'neural progenitor cell' divide asymmetrically through 1 or 2 rounds to produce cells that then differentiate into neurons, glial, epithelial cells. Lineage is determinate.

I have edited this proposed tasks to fit.

[github-actions]

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.