[bioontology-support] Problems with root concepts in a SKOS ontology

classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

[bioontology-support] Problems with root concepts in a SKOS ontology

Manuel Fiorelli
Dear BioPortal Support team

I've been doing some experiments with BioPortal using two private repositories, TESTONT (an OWL ontology) and TESTTES (a SKOS ontology).

Yesterday I submitted the attached RDF data to TESTTES (please note that the description of the concept scheme is split between two tags).

The file should conform to BioPortal constraints, such as the presence of a concept scheme and the use of skos:hasTopConcept.

However, the page that should retrieve the concepts informs me that there was an unspecified problem retrieving classes.


immagine.png

However, if I visit the URL corresponding to the detailed view of a concept, I can see the description of that concept together with the concept tree on the left side (note that the submission is a flat list of concepts).


immagine.png

PS
I should have asked  before; however, I took the opportunity to ask whether my use of private ontologies is good for experimentation, or whether there is a sandbox or other test system elsewhere. Indeed, I've just realized that there is no explicit delete button to remove the test ontologies.

Best regards
Manuel Fiorelli



--
Manuel Fiorelli

_______________________________________________
bioontology-support mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/bioontology-support

submission.rdf (5K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: [bioontology-support] Problems with root concepts in a SKOS ontology

jvendetti
Hello Manuel,

Apologies for a late response to your question. I’ve spent some time looking at the test SKOS ontology you submitted. Please find more information inline below.



On Feb 17, 2020, at 8:40 AM, Manuel Fiorelli <[hidden email]> wrote:

I've been doing some experiments with BioPortal using two private repositories, TESTONT (an OWL ontology) and TESTTES (a SKOS ontology).

Yesterday I submitted the attached RDF data to TESTTES (please note that the description of the concept scheme is split between two tags).

The file should conform to BioPortal constraints, such as the presence of a concept scheme and the use of skos:hasTopConcept.

However, the page that should retrieve the concepts informs me that there was an unspecified problem retrieving classes.




Whenever you navigate to the classes page for an ontology, BioPortal attempts to fetch the root nodes of the ontology in order to construct the class tree for display. For the ontology you submitted, BioPortal correctly detected individual classes during processing, which is evidenced by the following REST API call:




However, BioPortal was unable to calculate the root nodes of the ontology. The following REST API call to fetch root nodes returns an empty set:




This is why you see the error message in the user interface of "The page you are looking for wasn't found. Please try again."

I noticed that you declared a conceptScheme (“continents”), and placed a concept within this scheme using the skos:inScheme, and skos:topConceptOf constructs. This is of course perfectly valid, however currently BioPortal only detects root nodes using the skos:hasTopConcept construct. You can see that in our source code here:


In other words, if you declare your ontology using the hasTopConcept property - something like the following:

@prefix ex: <http://www.example.com/> .

ex:continents rdf:type skos:ConceptScheme;
  skos:hasTopConcept ex:Europe.


I believe that you would no longer get an error message in the BioPortal user interface because the system will be able to detect root classes. It’s a shortcoming of the current code base that we don’t detect roots using the other methodology, i.e., using inScheme and topConceptOf.



However, if I visit the URL corresponding to the detailed view of a concept, I can see the description of that concept together with the concept tree on the left side (note that the submission is a flat list of concepts).



This is a side effect of what I outlined above. The system can detect the individual concepts in your ontology, however it fails to detect the set of root classes it needs to properly construct the default page that displays the class tree as a whole.


PS I should have asked  before; however, I took the opportunity to ask whether my use of private ontologies is good for experimentation, or whether there is a sandbox or other test system elsewhere. Indeed, I've just realized that there is no explicit delete button to remove the test ontologies.


Private ontologies are fine. We don’t have the notion of a sandbox in our Stanford hosted BioPortal. If you want to delete any of your ontologies, feel free to contact us and we can do it on your behalf.

Kind regards,
Jennifer




_______________________________________________
bioontology-support mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/bioontology-support
Reply | Threaded
Open this post in threaded view
|

Re: [bioontology-support] Problems with root concepts in a SKOS ontology

Manuel Fiorelli
Hi Jennifer

thank you very much for the time that you spent on my request for support. I was aware of the limitation regarding the assignment of SKOS concepts to schemes; consequently, I added the "inverse" statements, which unfortunately were written at the end of the submission in a separate rdf:Description.

<rdf:Description rdf:about="http://example.org/conceptScheme_07199b34">
   <skos:hasTopConcept rdf:resource="http://example.org/c_48989b6a"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_e17f8367"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_4b2068df"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_4e52523a"/>
  <skos:hasTopConcept rdf:resource="http://example.org/c_968694e7"/>
</rdf:Description>

The motivation of this "strange" serialization is that it was generated automatically without enabling the pretty print routines. Do you think that this might explain why root classes aren't found.

In the meantime, I thank you for the confirmation about the use of private repositories

Regards
Manuel


Il giorno mar 3 mar 2020 alle ore 00:47 Jennifer Leigh Vendetti <[hidden email]> ha scritto:
Hello Manuel,

Apologies for a late response to your question. I’ve spent some time looking at the test SKOS ontology you submitted. Please find more information inline below.



On Feb 17, 2020, at 8:40 AM, Manuel Fiorelli <[hidden email]> wrote:

I've been doing some experiments with BioPortal using two private repositories, TESTONT (an OWL ontology) and TESTTES (a SKOS ontology).

Yesterday I submitted the attached RDF data to TESTTES (please note that the description of the concept scheme is split between two tags).

The file should conform to BioPortal constraints, such as the presence of a concept scheme and the use of skos:hasTopConcept.

However, the page that should retrieve the concepts informs me that there was an unspecified problem retrieving classes.




Whenever you navigate to the classes page for an ontology, BioPortal attempts to fetch the root nodes of the ontology in order to construct the class tree for display. For the ontology you submitted, BioPortal correctly detected individual classes during processing, which is evidenced by the following REST API call:




However, BioPortal was unable to calculate the root nodes of the ontology. The following REST API call to fetch root nodes returns an empty set:




This is why you see the error message in the user interface of "The page you are looking for wasn't found. Please try again."

I noticed that you declared a conceptScheme (“continents”), and placed a concept within this scheme using the skos:inScheme, and skos:topConceptOf constructs. This is of course perfectly valid, however currently BioPortal only detects root nodes using the skos:hasTopConcept construct. You can see that in our source code here:


In other words, if you declare your ontology using the hasTopConcept property - something like the following:

@prefix ex: <http://www.example.com/> .

ex:continents rdf:type skos:ConceptScheme;
  skos:hasTopConcept ex:Europe.


I believe that you would no longer get an error message in the BioPortal user interface because the system will be able to detect root classes. It’s a shortcoming of the current code base that we don’t detect roots using the other methodology, i.e., using inScheme and topConceptOf.



However, if I visit the URL corresponding to the detailed view of a concept, I can see the description of that concept together with the concept tree on the left side (note that the submission is a flat list of concepts).



This is a side effect of what I outlined above. The system can detect the individual concepts in your ontology, however it fails to detect the set of root classes it needs to properly construct the default page that displays the class tree as a whole.


PS I should have asked  before; however, I took the opportunity to ask whether my use of private ontologies is good for experimentation, or whether there is a sandbox or other test system elsewhere. Indeed, I've just realized that there is no explicit delete button to remove the test ontologies.


Private ontologies are fine. We don’t have the notion of a sandbox in our Stanford hosted BioPortal. If you want to delete any of your ontologies, feel free to contact us and we can do it on your behalf.

Kind regards,
Jennifer





--
Manuel Fiorelli

_______________________________________________
bioontology-support mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/bioontology-support
Reply | Threaded
Open this post in threaded view
|

Re: [bioontology-support] Problems with root concepts in a SKOS ontology

jvendetti
Hello Manuel,


On Mar 3, 2020, at 8:25 AM, Manuel Fiorelli <[hidden email]> wrote:

Hi Jennifer

thank you very much for the time that you spent on my request for support. I was aware of the limitation regarding the assignment of SKOS concepts to schemes; consequently, I added the "inverse" statements, which unfortunately were written at the end of the submission in a separate rdf:Description.

<rdf:Description rdf:about="http://example.org/conceptScheme_07199b34">
   <skos:hasTopConcept rdf:resource="http://example.org/c_48989b6a"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_e17f8367"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_4b2068df"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_4e52523a"/>
  <skos:hasTopConcept rdf:resource="http://example.org/c_968694e7"/>
</rdf:Description>

The motivation of this "strange" serialization is that it was generated automatically without enabling the pretty print routines. Do you think that this might explain why root classes aren't found.


When I investigated your problem report, I downloaded the file that was associated with your latest submission (the 6th file that you uploaded that is marked as version “1000” in the user interface). That source file doesn’t contain any of the hasTopConcept properties that you list above.

After reading this follow-up message from you, I downloaded one of your previous submissions (the 5th file that you uploaded). In that source file, I do see these hasTopConcept properties. It shouldn’t matter that they are listed at the end of your file in a separate rdf:Description. I went ahead and uploaded this source file to our staging environment (this is a private instance that we use internally for testing), and I can confirm that BioPortal properly detected the root classes and the 404 error no longer appears in the user interface:



 


Here is the REST call to retrieve the roots that no longer returns an empty set:




If you would like to upload a new version of your source file with these hasTopConcept properties, you should be able to successfully see the class tree in the user interface. One thing to note is that BioPortal has several caches in place for performance reasons. When you upload a new submission, the new data may not appear immediately.

Kind regards,
Jennifer



_______________________________________________
bioontology-support mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/bioontology-support
Reply | Threaded
Open this post in threaded view
|

Re: [bioontology-support] Problems with root concepts in a SKOS ontology

Manuel Fiorelli
Hi Jennifer

I made a new submission this afternoon, which should contain the triples with the predicate hasTopConcept.

I can fetch roots with the REST API:


but still the user interfaces produces an error. I hope that this is an issue with caching. I will try again in the next few days.

Regards
Manuel

Il giorno mer 4 mar 2020 alle ore 00:57 Jennifer Leigh Vendetti <[hidden email]> ha scritto:
Hello Manuel,


On Mar 3, 2020, at 8:25 AM, Manuel Fiorelli <[hidden email]> wrote:

Hi Jennifer

thank you very much for the time that you spent on my request for support. I was aware of the limitation regarding the assignment of SKOS concepts to schemes; consequently, I added the "inverse" statements, which unfortunately were written at the end of the submission in a separate rdf:Description.

<rdf:Description rdf:about="http://example.org/conceptScheme_07199b34">
   <skos:hasTopConcept rdf:resource="http://example.org/c_48989b6a"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_e17f8367"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_4b2068df"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_4e52523a"/>
  <skos:hasTopConcept rdf:resource="http://example.org/c_968694e7"/>
</rdf:Description>

The motivation of this "strange" serialization is that it was generated automatically without enabling the pretty print routines. Do you think that this might explain why root classes aren't found.


When I investigated your problem report, I downloaded the file that was associated with your latest submission (the 6th file that you uploaded that is marked as version “1000” in the user interface). That source file doesn’t contain any of the hasTopConcept properties that you list above.

After reading this follow-up message from you, I downloaded one of your previous submissions (the 5th file that you uploaded). In that source file, I do see these hasTopConcept properties. It shouldn’t matter that they are listed at the end of your file in a separate rdf:Description. I went ahead and uploaded this source file to our staging environment (this is a private instance that we use internally for testing), and I can confirm that BioPortal properly detected the root classes and the 404 error no longer appears in the user interface:



 


Here is the REST call to retrieve the roots that no longer returns an empty set:




If you would like to upload a new version of your source file with these hasTopConcept properties, you should be able to successfully see the class tree in the user interface. One thing to note is that BioPortal has several caches in place for performance reasons. When you upload a new submission, the new data may not appear immediately.

Kind regards,
Jennifer




--
Manuel Fiorelli

_______________________________________________
bioontology-support mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/bioontology-support
Reply | Threaded
Open this post in threaded view
|

Re: [bioontology-support] Problems with root concepts in a SKOS ontology

John Graybeal-2
Manuel, I'm seeing 5 countries as the classes and 5 properties, and several class mappings for each class.

If you use the Developer view on your browser (CMD-OPTION-I for MacOS Chrome) you can hold down the reload button to clear your local cache, that may do the trick.

john

On Mar 4, 2020, at 3:49 PM, Manuel Fiorelli <[hidden email]> wrote:

Hi Jennifer

I made a new submission this afternoon, which should contain the triples with the predicate hasTopConcept. 

I can fetch roots with the REST API:


but still the user interfaces produces an error. I hope that this is an issue with caching. I will try again in the next few days.

Regards
Manuel

Il giorno mer 4 mar 2020 alle ore 00:57 Jennifer Leigh Vendetti <[hidden email]> ha scritto:
Hello Manuel,


On Mar 3, 2020, at 8:25 AM, Manuel Fiorelli <[hidden email]> wrote:

Hi Jennifer

thank you very much for the time that you spent on my request for support. I was aware of the limitation regarding the assignment of SKOS concepts to schemes; consequently, I added the "inverse" statements, which unfortunately were written at the end of the submission in a separate rdf:Description.

<rdf:Description rdf:about="http://example.org/conceptScheme_07199b34">
   <skos:hasTopConcept rdf:resource="http://example.org/c_48989b6a"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_e17f8367"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_4b2068df"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_4e52523a"/>
  <skos:hasTopConcept rdf:resource="http://example.org/c_968694e7"/>
</rdf:Description>

The motivation of this "strange" serialization is that it was generated automatically without enabling the pretty print routines. Do you think that this might explain why root classes aren't found.


When I investigated your problem report, I downloaded the file that was associated with your latest submission (the 6th file that you uploaded that is marked as version “1000” in the user interface). That source file doesn’t contain any of the hasTopConcept properties that you list above.

After reading this follow-up message from you, I downloaded one of your previous submissions (the 5th file that you uploaded). In that source file, I do see these hasTopConcept properties. It shouldn’t matter that they are listed at the end of your file in a separate rdf:Description. I went ahead and uploaded this source file to our staging environment (this is a private instance that we use internally for testing), and I can confirm that BioPortal properly detected the root classes and the 404 error no longer appears in the user interface:



<Screenshot 2020-03-03 15.46.09.png> 


Here is the REST call to retrieve the roots that no longer returns an empty set:


<Screenshot 2020-03-03 15.53.59.png>


If you would like to upload a new version of your source file with these hasTopConcept properties, you should be able to successfully see the class tree in the user interface. One thing to note is that BioPortal has several caches in place for performance reasons. When you upload a new submission, the new data may not appear immediately.

Kind regards,
Jennifer




-- 
Manuel Fiorelli
_______________________________________________
bioontology-support mailing list
[hidden email]
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  | ORCID  0000-0001-6875-5360





Begin forwarded message:

From: Manuel Fiorelli <[hidden email]>

Subject: Re: [bioontology-support] Problems with root concepts in a SKOS ontology

Date: March 4, 2020 at 3:49:46 PM PST

To: Jennifer Leigh Vendetti <[hidden email]>

Cc: "[hidden email]" <[hidden email]>


Hi Jennifer

I made a new submission this afternoon, which should contain the triples with the predicate hasTopConcept.

I can fetch roots with the REST API:


but still the user interfaces produces an error. I hope that this is an issue with caching. I will try again in the next few days.

Regards
Manuel

Il giorno mer 4 mar 2020 alle ore 00:57 Jennifer Leigh Vendetti <[hidden email]> ha scritto:
Hello Manuel,


On Mar 3, 2020, at 8:25 AM, Manuel Fiorelli <[hidden email]> wrote:

Hi Jennifer

thank you very much for the time that you spent on my request for support. I was aware of the limitation regarding the assignment of SKOS concepts to schemes; consequently, I added the "inverse" statements, which unfortunately were written at the end of the submission in a separate rdf:Description.

<rdf:Description rdf:about="http://example.org/conceptScheme_07199b34">
   <skos:hasTopConcept rdf:resource="http://example.org/c_48989b6a"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_e17f8367"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_4b2068df"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_4e52523a"/>
  <skos:hasTopConcept rdf:resource="http://example.org/c_968694e7"/>
</rdf:Description>

The motivation of this "strange" serialization is that it was generated automatically without enabling the pretty print routines. Do you think that this might explain why root classes aren't found.


When I investigated your problem report, I downloaded the file that was associated with your latest submission (the 6th file that you uploaded that is marked as version “1000” in the user interface). That source file doesn’t contain any of the hasTopConcept properties that you list above.

After reading this follow-up message from you, I downloaded one of your previous submissions (the 5th file that you uploaded). In that source file, I do see these hasTopConcept properties. It shouldn’t matter that they are listed at the end of your file in a separate rdf:Description. I went ahead and uploaded this source file to our staging environment (this is a private instance that we use internally for testing), and I can confirm that BioPortal properly detected the root classes and the 404 error no longer appears in the user interface:



 


Here is the REST call to retrieve the roots that no longer returns an empty set:




If you would like to upload a new version of your source file with these hasTopConcept properties, you should be able to successfully see the class tree in the user interface. One thing to note is that BioPortal has several caches in place for performance reasons. When you upload a new submission, the new data may not appear immediately.

Kind regards,
Jennifer




--
Manuel Fiorelli
_______________________________________________
bioontology-support mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/bioontology-support

_______________________________________________
bioontology-support mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/bioontology-support
Reply | Threaded
Open this post in threaded view
|

Re: [bioontology-support] Problems with root concepts in a SKOS ontology

Manuel Fiorelli
Hi John

I logged in BioPortal and now I can see my thesaurus. However, I don't know whether it was a server-side or client-side cache.

I would like to thank you and Jennifer for the support.

Regards
Manuel


Il giorno gio 5 mar 2020 alle ore 02:37 John Graybeal <[hidden email]> ha scritto:
Manuel, I'm seeing 5 countries as the classes and 5 properties, and several class mappings for each class.

If you use the Developer view on your browser (CMD-OPTION-I for MacOS Chrome) you can hold down the reload button to clear your local cache, that may do the trick.

john

On Mar 4, 2020, at 3:49 PM, Manuel Fiorelli <[hidden email]> wrote:

Hi Jennifer

I made a new submission this afternoon, which should contain the triples with the predicate hasTopConcept. 

I can fetch roots with the REST API:


but still the user interfaces produces an error. I hope that this is an issue with caching. I will try again in the next few days.

Regards
Manuel

Il giorno mer 4 mar 2020 alle ore 00:57 Jennifer Leigh Vendetti <[hidden email]> ha scritto:
Hello Manuel,


On Mar 3, 2020, at 8:25 AM, Manuel Fiorelli <[hidden email]> wrote:

Hi Jennifer

thank you very much for the time that you spent on my request for support. I was aware of the limitation regarding the assignment of SKOS concepts to schemes; consequently, I added the "inverse" statements, which unfortunately were written at the end of the submission in a separate rdf:Description.

<rdf:Description rdf:about="http://example.org/conceptScheme_07199b34">
   <skos:hasTopConcept rdf:resource="http://example.org/c_48989b6a"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_e17f8367"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_4b2068df"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_4e52523a"/>
  <skos:hasTopConcept rdf:resource="http://example.org/c_968694e7"/>
</rdf:Description>

The motivation of this "strange" serialization is that it was generated automatically without enabling the pretty print routines. Do you think that this might explain why root classes aren't found.


When I investigated your problem report, I downloaded the file that was associated with your latest submission (the 6th file that you uploaded that is marked as version “1000” in the user interface). That source file doesn’t contain any of the hasTopConcept properties that you list above.

After reading this follow-up message from you, I downloaded one of your previous submissions (the 5th file that you uploaded). In that source file, I do see these hasTopConcept properties. It shouldn’t matter that they are listed at the end of your file in a separate rdf:Description. I went ahead and uploaded this source file to our staging environment (this is a private instance that we use internally for testing), and I can confirm that BioPortal properly detected the root classes and the 404 error no longer appears in the user interface:



<Screenshot 2020-03-03 15.46.09.png> 


Here is the REST call to retrieve the roots that no longer returns an empty set:


<Screenshot 2020-03-03 15.53.59.png>


If you would like to upload a new version of your source file with these hasTopConcept properties, you should be able to successfully see the class tree in the user interface. One thing to note is that BioPortal has several caches in place for performance reasons. When you upload a new submission, the new data may not appear immediately.

Kind regards,
Jennifer




-- 
Manuel Fiorelli
_______________________________________________
bioontology-support mailing list
[hidden email]
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  | ORCID  0000-0001-6875-5360





Begin forwarded message:

From: Manuel Fiorelli <[hidden email]>

Subject: Re: [bioontology-support] Problems with root concepts in a SKOS ontology

Date: March 4, 2020 at 3:49:46 PM PST

To: Jennifer Leigh Vendetti <[hidden email]>

Cc: "[hidden email]" <[hidden email]>


Hi Jennifer

I made a new submission this afternoon, which should contain the triples with the predicate hasTopConcept.

I can fetch roots with the REST API:


but still the user interfaces produces an error. I hope that this is an issue with caching. I will try again in the next few days.

Regards
Manuel

Il giorno mer 4 mar 2020 alle ore 00:57 Jennifer Leigh Vendetti <[hidden email]> ha scritto:
Hello Manuel,


On Mar 3, 2020, at 8:25 AM, Manuel Fiorelli <[hidden email]> wrote:

Hi Jennifer

thank you very much for the time that you spent on my request for support. I was aware of the limitation regarding the assignment of SKOS concepts to schemes; consequently, I added the "inverse" statements, which unfortunately were written at the end of the submission in a separate rdf:Description.

<rdf:Description rdf:about="http://example.org/conceptScheme_07199b34">
   <skos:hasTopConcept rdf:resource="http://example.org/c_48989b6a"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_e17f8367"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_4b2068df"/>
   <skos:hasTopConcept rdf:resource="http://example.org/c_4e52523a"/>
  <skos:hasTopConcept rdf:resource="http://example.org/c_968694e7"/>
</rdf:Description>

The motivation of this "strange" serialization is that it was generated automatically without enabling the pretty print routines. Do you think that this might explain why root classes aren't found.


When I investigated your problem report, I downloaded the file that was associated with your latest submission (the 6th file that you uploaded that is marked as version “1000” in the user interface). That source file doesn’t contain any of the hasTopConcept properties that you list above.

After reading this follow-up message from you, I downloaded one of your previous submissions (the 5th file that you uploaded). In that source file, I do see these hasTopConcept properties. It shouldn’t matter that they are listed at the end of your file in a separate rdf:Description. I went ahead and uploaded this source file to our staging environment (this is a private instance that we use internally for testing), and I can confirm that BioPortal properly detected the root classes and the 404 error no longer appears in the user interface:



 


Here is the REST call to retrieve the roots that no longer returns an empty set:




If you would like to upload a new version of your source file with these hasTopConcept properties, you should be able to successfully see the class tree in the user interface. One thing to note is that BioPortal has several caches in place for performance reasons. When you upload a new submission, the new data may not appear immediately.

Kind regards,
Jennifer




--
Manuel Fiorelli
_______________________________________________
bioontology-support mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/bioontology-support


--
Manuel Fiorelli

_______________________________________________
bioontology-support mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/bioontology-support