Is an apical dendrite a part of apical part of cell?

[gocentral]

Or does the current GO term refer really to the apical part of the cell soma in the case of neurons or other cells with projections? http://amigo.geneontology.org/cgi-bin/amigo/term\_details?term=GO:0045177

Reported by: alanr

Original Ticket: geneontology/ontology-requests/8580

[gocentral]

• assigned_to: nobody --> paolaroncaglia

Original comment by: paolaroncaglia

[gocentral]

• labels: 397373 --> neurobiology

Original comment by: paolaroncaglia

[gocentral]

Hi Alan, To keep it simple, I'd say:

An apical dendrite is a dendrite that emerges from the apex of a pyramidal cell. (http://en.wikipedia.org/wiki/Apical\_dendrite)

And a pyramidal cell is a neuron that has a pyramid-shaped soma with the apex and an apical dendrite pointed toward the pial surface, and other dendrites and an axon emerging from the base. (Cell Ontology, CL:0000598)

A pyramidal cell can be considered a polarized cell (because it faces the pial surface on one end), and it can therefore be said to have an apical part, which is the one that faces the pial surface;

But the apical dendrite, being a dendrite, is_a neuron projection, which is_a cell projection. These terms don't have relationships with other cell parts, including apical part of cell, which I guess we all agree on.

Therefore, I wouldn't make "apical dendrite" part_of apical part of cell (GO:0045177). I would, however, create new GO terms for "apical dendrite" and "basal dendrite" (synonym: basilar dendrite). The latter also emerges from pyramidal cells.

Let me know if you disagree with any of the above, otherwise I'll go ahead. Thanks,

Paola

Original comment by: paolaroncaglia

[gocentral]

Makes sense:

I would, however, create new GO terms for "apical dendrite" and "basal dendrite" (synonym: basilar dendrite). The latter also emerges from pyramidal cells.

I think you want to make these part_of pyramidal cell, and perhaps overlaps with apical part of pyramidal cell/basal part of pyramidal call

You may also want to have apical pole of pyramidal cell following http://neurolex.org/wiki/Category:Apical\_Pole\_Of\_Neuron

However, there is a question of whether the restriction to pyramidal cells is too broad. See, for instance, http://neurolex.org/wiki/Category:Olfactory\_bulb\_\(accessory\)\_granule\_cell

I will send an email to the INCF mailing list cc you to discuss this.

Original comment by: alanr

[gocentral]

Alan and Paola,

I don't think the GO should add a part_of relationship between "apical dendrite" or "basal dendrite" and "pyramidal cell". I haven't seen any proof in this thread that these projection types only occur in pyramidal neurons and not other types of neurons, and without a wide-ranging survey of the literature and the experts I would be very reluctant to encode this restriction in the GO. I also don't think we should rely on Wikipedia for a definition resolution of this somewhat detailed and subtle point in neuroscience.

A has_part relationship for the relevant CL term seems much more appropriate.

I do agree that neurons do not necessarily have apical and basel parts analogous to other cell types as the cellular and tissue morphology is clearly quite different. This too seems like a point for expert input.

-- Alex

Original comment by: addiehl

[gocentral]

I would like to add a part_of only (x or y or z) IF we get a response from the experts that this is the case. This is the sort of constraint that leverages well in finding errors. If the experts say that we shouldn't say anything of the sort, then we don't need to do this.

OTOH the other alternative I offered, which is to restrict by morphology can equally well be leverages - we can say that a cell soma can only have one global shape quality and then assert the soma shape qualities. If we then say that apical dendrites are part of only cells that have the right morphology then we will also be able to easily see erroneous assertions.

Such restrictions don't only help in QC, but also better document what we mean by our terms.

Original comment by: alanr

[gocentral]

Apologies for the delay in getting back to this; I've now resolved this in conjunction with the ongoing integration of SAO terms in GO as cellular components.

I've added the following terms, all as is_a children of 'dendrite':

GO:0097440 apical dendrite GO:0097441 basilar dendrite GO:0097442 CA3 pyramidal cell dendrite

I have not added part_of links to these terms because, based on previous email discussions, not all neurons have apical/basal dendrites, but many do and it's difficult to restrict the terms to some neuron types only.

Thanks, Paola

Original comment by: paolaroncaglia

[gocentral]

• status: open --> closed-fixed

Original comment by: paolaroncaglia