Hi,
Could the usage of dash in labels be homogenized (or is there a reason for having differences)? For example "T cell receptor co-receptor CD8" has_part some 'T-cell surface glycoprotein [..]'
Having or not a dash between "T" and "cell" arbitrarily, coupled with underscore or not in relations makes it very hard to write a SPARQL query; even more so if I am trying to use the Protege auto-completion feature.
Thanks,
Melanie
Hi Melanie,
I'm not sure why this is a request to the CL tracker as the terms you reference are Protein Ontology terms.
As for the Cell Ontology, the fact that CL terms for T cells and B cells lack hyphens was not at all an arbitrary decision. Years ago I checked Pubmed and saw that "T cell" and "B cell" were the most common usage by biologists, more common than T-cell, T lymphocyte, and T-lymphocyte, and similarly for "B cell" and its variants. Thus CL terms referring to types of T cells or B cell use the hyphenless "T cell" or "B cell" as part of the term name, and most of these terms have exact synonyms using T-cell, T lymphocyte, or T-lymphocyte, etc. That's a lot of synonyms, and we may have missed a few terms here and there; I haven't checked lately. The participants at the NIAID Cell Ontology workshop in 2008 (mostly immunologists) agreed with this editorial decision.
PRO on the other hand, utilizes UniProt names for their proteins, and UniProt has not been consistent in its usage of T cell, T-cell, etc. Nor does PRO provide the wealth of synonyms that the CL provides.
I see no reason to change the CL as we have tried to be very consistent in our non-use of the hyphen here, based as I say on the usage of actual biologists, and very pro-active in providing exact synonyms. Tools that search terms via synonyms as well as term names should have no problem finding appropriate terms using T-cell or T-lymphocyte, etc. Such tools include AmiGO and OBO-Edit, although apparently not Protege. Perhaps this could be a request to the Protege developers.
Thanks,
Alex
Thanks Alex; sorry I didn't realize those were PRO terms. You seem to have given great thought to the issue; do you think this is something the CL developers would be willing to ask PRO, in the interest of harmonization between resources?
I will check with the Protege team if there is an option to add other annotations properties to the auto-completion feature.
GO and CL always use "T cell" as primary label.
I believe this request is intended for PRO?
On 2 Dec 2014, at 13:06, Nicole Vasilevsky wrote:
Related
Cell Ontology (CL) requests:
#197That is what Alex had mentioned yes, and I did talk to Darren Natale and Cathy Wu at one of the Buffalo meetings and they had agreed on homogenizing and at least add other forms as synonyms. I’m not sure whether this did or not happen, but we worked around it in our application.
Cheers,
Melanie
On Dec 2, 2014, at 10:58 PM, Chris Mungall cmungall@users.sf.net wrote:
Related
Cell Ontology (CL) requests:
#197