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] Ask help

Jennifer Leigh Vendetti vendetti at stanford.edu
Fri Sep 9 15:43:17 PDT 2016


Hello Hanfei,


On Sep 7, 2016, at 8:07 AM, 包含飞 <hanfeib at gmail.com<mailto:hanfeib at gmail.com>> wrote:

Hi Jennifer,

I have met a problem in importing ontology.

That is that

I have developed two new ontologies, one is about AIDS clinics

 (see http://bioportal.bioontology.org/ontologies/AIDSCLIN)

 and another one is about human gene codons

(see http://bioportal.bioontology.org/ontologies/HCODONONT/).

The two new ontologies need to import the main ontology

(see https://bioportal.bioontology.org/ontologies/HIVO004).

It seems the two new ontologies at web sites are not successful to import the main ontology(see Pic. 1) , unlike they are located in my local computer file folds(see Pic. 2).

I would ask you if you can help me checking out the exact address of my main ontology at BioPortal. Maybe it will be helpful for a valid import of ontology in the way from web to web.


I think there is a misunderstanding here in terms of the functionality that BioPortal provides.  BioPortal allows you to upload an ontology and we then provide a Web interface and a REST endpoint for viewing / retrieving ontology data such as classes, ontology metadata, etc.  However, we don’t provide a canonical URL where your ontology source file resides that you could use as part of an owl:import statement.  In other words, you can’t do this (from the source code of your HCODONONT ontology):

<owl:imports rdf:resource="http://bioportal/bioontology.org/ontologies/HIVO0004”/>

Opening both of your ontologies (AIDSCLIN and HCODONONT) in Protege displays error messages about a missing import.  (I’m referring to the most recent version of both, downloaded from BioPortal).

You have a couple of options for getting imports to work properly:

1).  If you want to use a web address in your import statement, you could publish your main ontology source file (HIVO004) at some URL that you have control / have access to.  You could then update the import statement in AIDSCLIN and HCODONONT to use that URL.

2).  If you’re unable to publish your main ontology file at some URL, you could edit your importing ontologies to import from a local file (you could use Protege for this step).  Then, you would need to upload new versions to BioPortal in the form of ZIP files that contain both the importing and imported ontology files.  It’s not particularly well documented, but BioPortal does allow you to upload ZIP files that contain multiple ontology files.  If you want to do this, you could refer to this post for instructions:

http://ncbo-support.2288202.n4.nabble.com/bioontology-support-Multi-file-ontology-td4363558.html#a4363572

… or let us know if you have issues and we can try to help further.

Best,
Jennifer

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.stanford.edu/pipermail/bioontology-support/attachments/20160909/684888a3/attachment.html>


More information about the bioontology-support mailing list