Search Mailing List Archives


Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

[bioontology-support] Help with Submission

Samson Tu swt at stanford.edu
Thu Jan 25 20:03:50 PST 2018


Hi,

I’d endorse John’s advocacy of “mak[ing] the ontology correct (i.e., do not misrepresent partOf relationships as subclasses.)." Where you have “part-of” relation in your domain, don’t modeled it as “subclass_of.” Perhaps you want to create a third version where “part-of” is never modeled as “subclass_of”.

The next version of Protege (which may become available in February) will display relationships other than subclass-of. In the current version of Protege, there is an Individual Hierarchy plugin that you can install that can display object-property assertions as hierarchies at the individual level. It may not be what you need, but I am mentioning it just in case.

With best regards,
Samson

On Jan 25, 2018, at 3:12 PM, Pan, Huaqin (Helen) <hpan at rti.org<mailto:hpan at rti.org>> wrote:

Hi Jennifer and John,

Thank you both for your advice! I will submit both versions as private first, and choose the “correct” one to make public.

John, I appreciate your suggestion in the last paragraph. Yes, I didn’t do that (have both “part-of” AND “subclass”). In version 1, I don’t have “part-of”. In version 2 where I have “part-of”, it’s either “part-of” or “subclass_of”, not BOTH.

Best regards,
Helen

From: John Graybeal [mailto:jgraybeal at stanford.edu]
Sent: Wednesday, January 24, 2018 10:59 PM
To: Jennifer Leigh Vendetti <vendetti at stanford.edu<mailto:vendetti at stanford.edu>>
Cc: Pan, Huaqin (Helen) <hpan at rti.org<mailto:hpan at rti.org>>; NCBO User Support <support at bioontology.org<mailto:support at bioontology.org>>
Subject: Re: [bioontology-support] Help with Submission

I might add a few thoughts to Jennifer's summary.

My bias in such things is to make the ontology correct (i.e., do not misrepresent partOf relationships as subclasses, because it messes up any inferencing that people may want to do (and maybe quite important research, too). It's better to 'force' the tools to improve than to create improper data artifacts, even if it is painful.

Re the visualization, I understand the Protege team may be working on an improved visual presentation of tree hierarchies, partly to address this issue—I just saw a note to that effect in another list, but haven't asked the team. If that's the case, we would definitely see if leveraging that code for BioPortal could be possible, though it wouldn't happen quickly.

Regarding the proposed idea to publish them privately at first—since you don't know which one you'll make public, you won't know which one to give the correct acronym. I'd give them both 'fake' acronyms, then republish the one you like publicly under the correct acronym. (We could delete the private ones afterwards, that would be best for BioPortal.)

And finally, a really bad suggestion given my bias above: if you basically _added_ a subclass relation for each partOf relation, if the ontology parsed (not sure if it could), you'd have both worlds—a hierarchical view and still being able to see the partOf relations. But I really hope you don't do that.

John


On Jan 24, 2018, at 5:18 PM, Jennifer Leigh Vendetti <vendetti at stanford.edu<mailto:vendetti at stanford.edu>> wrote:

Hi Helen,



On Jan 22, 2018, at 1:59 PM, Pan, Huaqin (Helen) <hpan at rti.org<mailto:hpan at rti.org>> wrote:

Hi Jennifer,

I am working on another project that developed an ontology that I will submit soon. It’s currently available at our project site, https://lungmap.net/breath-ontology-browser/.

Here is my questions. Due to the fact that owl/Protégé display the hierarchical relation only for “subclass-of”, not “part-of”, I have developed 2 versions of the ontology in .owl for each ontology, and I need your guidance on which version to submit. I have seen both on BioPortal.


My expertise is on the software development side of things with regard to BioPortal. It’s difficult for me to comment on which version of your ontology the biomedical community might find more useful.





  *   Version 1 with exclusive “subclass_of” – Pros: display the hierarchical relationships among the classes, Cons: lacks the biological distinction of “part-of” and “subclass_of”
  *   Version 2 with mix of “part-of” and “subclass_of” – Cons: lacks the display of hierarchical relationships among the classes, Pros: captures the biological distinction of “part-of” and “subclass_of”


We used version 2 (part-of and subclass-of) for the project site, with some rules to enable the display properly for “part-of”.


One option I can suggest would be for you to create two separate ontology entries in BioPortal for your versions 1 and 2 above. You could initially make the viewing restriction “private”, which would give you the opportunity to view them and decide which seems better for your needs / use cases. After you’ve made your decision, we could remove whichever one you decide not to publish, and change the viewing restriction of your preferred ontology to “public”.

Kind regards,
Jennifer

_______________________________________________
bioontology-support mailing list
bioontology-support at lists.stanford.edu<mailto:bioontology-support at lists.stanford.edu>
https://mailman.stanford.edu/mailman/listinfo/bioontology-support

========================
John Graybeal
Technical Program Manager
Center for Expanded Data Annotation and Retrieval /+/ NCBO BioPortal
Stanford Center for Biomedical Informatics Research
650-736-1632

_______________________________________________
bioontology-support mailing list
bioontology-support at lists.stanford.edu<mailto:bioontology-support at lists.stanford.edu>
https://mailman.stanford.edu/mailman/listinfo/bioontology-support

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.stanford.edu/pipermail/bioontology-support/attachments/20180126/a4d5ea15/attachment-0001.html>


More information about the bioontology-support mailing list