[Ontology-editors] [Obo-cell-type] help wanted getting CL and GO consistent

[Chris Mungall]

On Oct 7, 2008, at 2:26 AM, David Sutherland wrote:

> On Mon, 2008-10-06 at 14:39 -0700, Chris Mungall wrote:
>> We have a set of computable cross-product definitions for a subset of
>> GO-biological process terms referencing cell types. See:
>>
>> 	http://wiki.geneontology.org/index.php/XP:biological_process_xp_cell
>>
>> Using the OBO-Edit reasoner we can find implied links in GO. The
>> policy in GO is to redundantly assert links that can be implied by  
>> the
>> reasoner through xp defs.
>>
>> For example, CL says:
>>
>> * oenocyte is columnar/cuboidal epithelial cell
>>
>> The xp defs say:
>>
>> * oenocyte differentiation = cell differentiation that
>> results_in_acquisition_of_features_of oenocyte
>> * columnar/cuboidal epithelial cell differentiation = cell
>> differentiation that results_in_acquisition_of_features_of columnar/
>> cuboidal epithelial cell
>>
>> Given this, a reasoner will infer:
>>
>> * oenocyte differentiation is_a columnar/cuboidal epithelial cell
>> differentiation
>>
>> This link is not present in GO
>>
>> There is a summary of these results here:
>>
>> https://sourceforge.net/tracker/index.php?func=detail&aid=1940574&group_id=36855&atid=440764
>>
>> - Many of these pertain to immunology; Alex suggested waiting on CL
>> revisions based on recent immunology work. This seems reasonable -
>> provided this is reasonably imminent. If not, I think we should just
>> go in and make things consistent.
>>
>> - The non-immunology ones seem fairly straightforward cases of one
>> ontology being right and the other being wrong.
>>
>> - David S : can you comment on oenocyte?
>
> This is wrong.  Oenocytes are *not* epithelial.  They are large cells
> which sit in clumps under the epidermis.  Function is not well
> characterised, but I believe there is some evidence that the adult
> oenocytes have some function in pheromone release.

Great, we're getting somewhere!

I will move oenocyte to be an is_a child of cell for now. We can find  
better is_a parentage for it later, the important thing for now is to  
be not wrong.

The current structure is somewhat curious. Here is the is_a hierarchy:

      is_a CL:0000066 ! epithelial cell
       is_a CL:0000075 ! columnar/cuboidal epithelial cell
        is_a CL:0000463 ! epidermal cell (sensu Insecta)  
[develops_from: epidermoblast]
         is_a CL:0000487 ! oenocyte ***

oenocyte has no is_a sibling, so I will go ahead and obsolete CL: 
0000463 ! epidermal cell (sensu Insecta)

This term seems to exist purely to provide developmental lineage to  
epidermoblast. I can retain this by making a direct develops_from link  
from oenocyte to epidermoblast -- but then we would be making a  
stronger statement than is present in fly_anatomy, which provides no  
developmental lineage for oenocyte, even over part_of. I will thus err  
on the side of caution for now.

This was CL previously:


-------------- next part --------------
A non-text attachment was scrubbed...
Name: oenocyte.png
Type: image/png
Size: 21574 bytes
Desc: not available
URL: <http://fafner.stanford.edu/pipermail/ontology-editors/attachments/20081007/877aa13f/attachment-0001.png>
-------------- next part --------------





> David
>>
>> I volunteer to make required changes to CL - just send me a list of
>> what needs fixed
>>
>> I also volunteer to fix GO too (I can check in directly or just send
>> someone an updated file to check in). There's no reason to do this by
>> hand when it can be done with a few clicks in OE. I can also give
>> another demo of how this can be done in OE if people are interested.
>>
>> Cheers
>> Chris
>>
>> -------------------------------------------------------------------------
>> This SF.Net email is sponsored by the Moblin Your Move Developer's  
>> challenge
>> Build the coolest Linux based applications with Moblin SDK & win  
>> great prizes
>> Grand prize is a trip for two to an Open Source event anywhere in  
>> the world
>> http://moblin-contest.org/redirect.php?banner_id=100&url=/
>> _______________________________________________
>> Obo-cell-type mailing list
>> Obo-cell-type at lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/obo-cell-type
>
>