Loading SNOMED Clinical Findings ontology into Protege

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

Loading SNOMED Clinical Findings ontology into Protege

Michael DeBellis-2
I realize this is a question more for BioPortal but I recall that some researchers posted to this list a while ago who where were working on SNOMED and last time I asked a question to the Bioportal site I didn't get a very detailed answer (although I'll try them again as well). BTW, if there is a user group or list specifically for SNOMED any links would be appreciated. 

I tried loading the OWL file that can be found here: http://bioportal.bioontology.org/ontologies/SNOMED_CF  (the file at the bottom of that page dated 5/7/2014). The file takes a very long time to load but ultimately fails due to an error that I think indicates Java ran out of virtual memory (I'll post the log info below). A while back I was trying to load a large ontology and I needed to increase the parameters on my ini file. My current ini file is:

-Xms200M
-Xmx2000M
-Xss16M

I'm running Windows 10 Home version on a desktop with 12GB of RAM.

My questions are: 
1)  Is it feasible to load this whole ontology on my kind of desktop by further editing my ini file? I could also purchase more RAM if that would help. Any recommendations as to what I should change the ini parameters to to make it work? I have to confess I don't understand what these parameters are doing, I just know someone suggested I increase one of them and I did and it worked on other large ontologies up until now.

2) If the ontology is just too large is there a way to extract some parts of it? I would like to do that anyway because I don't really need the whole ontology. I tried using the "File>Open from URL" option and gave it the ID for one high level class from that ontology (the disease class: http://ihtsdo.org/snomedct/clinicalFinding#64572001) I was hoping it might just load that class and all it's subclasses but after churning for a while I got an error that Protege couldn't parse the ontology. 

3) Are there other ways to get this to work such as use a triple store? Any suggestions as to tutorials or other info about what or how to use in this case?

Here is the log file below. I cleared the warning messages that I get during a normal startup so these are just the messages after trying to load that ontology, this was a fresh start so no issues of other ontologies taking up space. Any feedback would be appreciated:

------------------------------ Log file ----------------------------------------------

   INFO  12:47:31  OWL API Version: 4.5.9.2019-02-01T07:24:44Z
   INFO  12:47:34  ------------------------------- Loading Ontology -------------------------------
   INFO  12:47:34  Loading ontology from file:/C:/Users/Michael%20DeBellis/Desktop/Ontologies/Neha/snomed_cf.owl/snomed_cf.owl
   INFO  13:06:10  ---------------------------- Disposing of Workspace ----------------------------
   INFO  13:06:10  Finished loading file:/C:/Users/Michael%20DeBellis/Desktop/Ontologies/Neha/snomed_cf.owl/snomed_cf.owl
   INFO  13:06:10  Saved tab state for 'Object properties' tab
   INFO  13:06:10  Saved tab state for 'DL Query' tab
   INFO  13:06:10  Saved tab state for 'Entities' tab
   INFO  13:06:10  Saved tab state for 'Classes' tab
   INFO  13:06:10  Saved tab state for 'SWRLTab' tab
   INFO  13:06:10  Saved tab state for 'Data properties' tab
   INFO  13:06:10  Saved tab state for 'Active ontology' tab
   INFO  13:06:10  Saved tab state for 'SQWRLTab' tab
   INFO  13:06:10  Saved tab state for 'Individuals by class' tab
   INFO  13:06:10  Saved tab state for 'SPARQL Query' tab
   INFO  13:06:10  Saved workspace
   INFO  13:06:10  Disposed of 'Object properties' tab
   INFO  13:06:10  Disposed of 'DL Query' tab
   INFO  13:06:10  Disposed of 'Entities' tab
   INFO  13:06:10  Disposed of 'Classes' tab
   INFO  13:06:10  Disposed of 'SWRLTab' tab
   INFO  13:06:10  Disposed of 'Data properties' tab
   INFO  13:06:10  Disposed of 'Active ontology' tab
   INFO  13:06:10  Disposed of 'SQWRLTab' tab
   INFO  13:06:10  Disposed of 'Individuals by class' tab
   INFO  13:06:10  Disposed of 'SPARQL Query' tab
   INFO  13:06:10  Disposed of workspace
   INFO  13:06:10  
  ERROR  13:06:10  Uncaught Exception in thread 'AWT-EventQueue-0'
java.lang.OutOfMemoryError: Java heap space

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

Re: Loading SNOMED Clinical Findings ontology into Protege

samsontu
Hi,

I have no problem loading a downloaded copy of SNOMED CF in Protege 5.5 that has the setting of Xmx3000M.

SNOMED CF is the extracted “findings” hierarchy of SNOMED CT (as of 2014).

I have no problem loading the 2019 version of SNOMED CT with the Xmx5000M setting.

I am using a 2013 MacBook Pro.

The earlier discussion was about loading SNOMED CT+LOINC+RxNORM.

With best regards,
Samson


On Nov 27, 2019, at 9:34 AM, Michael DeBellis <[hidden email]> wrote:

I realize this is a question more for BioPortal but I recall that some researchers posted to this list a while ago who where were working on SNOMED and last time I asked a question to the Bioportal site I didn't get a very detailed answer (although I'll try them again as well). BTW, if there is a user group or list specifically for SNOMED any links would be appreciated. 

I tried loading the OWL file that can be found here: http://bioportal.bioontology.org/ontologies/SNOMED_CF  (the file at the bottom of that page dated 5/7/2014). The file takes a very long time to load but ultimately fails due to an error that I think indicates Java ran out of virtual memory (I'll post the log info below). A while back I was trying to load a large ontology and I needed to increase the parameters on my ini file. My current ini file is:

-Xms200M
-Xmx2000M
-Xss16M

I'm running Windows 10 Home version on a desktop with 12GB of RAM.

My questions are: 
1)  Is it feasible to load this whole ontology on my kind of desktop by further editing my ini file? I could also purchase more RAM if that would help. Any recommendations as to what I should change the ini parameters to to make it work? I have to confess I don't understand what these parameters are doing, I just know someone suggested I increase one of them and I did and it worked on other large ontologies up until now.

2) If the ontology is just too large is there a way to extract some parts of it? I would like to do that anyway because I don't really need the whole ontology. I tried using the "File>Open from URL" option and gave it the ID for one high level class from that ontology (the disease class: http://ihtsdo.org/snomedct/clinicalFinding#64572001) I was hoping it might just load that class and all it's subclasses but after churning for a while I got an error that Protege couldn't parse the ontology. 

3) Are there other ways to get this to work such as use a triple store? Any suggestions as to tutorials or other info about what or how to use in this case?

Here is the log file below. I cleared the warning messages that I get during a normal startup so these are just the messages after trying to load that ontology, this was a fresh start so no issues of other ontologies taking up space. Any feedback would be appreciated:

------------------------------ Log file ----------------------------------------------

   INFO  12:47:31  OWL API Version: 4.5.9.2019-02-01T07:24:44Z
   INFO  12:47:34  ------------------------------- Loading Ontology -------------------------------
   INFO  12:47:34  Loading ontology from file:/C:/Users/Michael%20DeBellis/Desktop/Ontologies/Neha/snomed_cf.owl/snomed_cf.owl
   INFO  13:06:10  ---------------------------- Disposing of Workspace ----------------------------
   INFO  13:06:10  Finished loading file:/C:/Users/Michael%20DeBellis/Desktop/Ontologies/Neha/snomed_cf.owl/snomed_cf.owl
   INFO  13:06:10  Saved tab state for 'Object properties' tab
   INFO  13:06:10  Saved tab state for 'DL Query' tab
   INFO  13:06:10  Saved tab state for 'Entities' tab
   INFO  13:06:10  Saved tab state for 'Classes' tab
   INFO  13:06:10  Saved tab state for 'SWRLTab' tab
   INFO  13:06:10  Saved tab state for 'Data properties' tab
   INFO  13:06:10  Saved tab state for 'Active ontology' tab
   INFO  13:06:10  Saved tab state for 'SQWRLTab' tab
   INFO  13:06:10  Saved tab state for 'Individuals by class' tab
   INFO  13:06:10  Saved tab state for 'SPARQL Query' tab
   INFO  13:06:10  Saved workspace
   INFO  13:06:10  Disposed of 'Object properties' tab
   INFO  13:06:10  Disposed of 'DL Query' tab
   INFO  13:06:10  Disposed of 'Entities' tab
   INFO  13:06:10  Disposed of 'Classes' tab
   INFO  13:06:10  Disposed of 'SWRLTab' tab
   INFO  13:06:10  Disposed of 'Data properties' tab
   INFO  13:06:10  Disposed of 'Active ontology' tab
   INFO  13:06:10  Disposed of 'SQWRLTab' tab
   INFO  13:06:10  Disposed of 'Individuals by class' tab
   INFO  13:06:10  Disposed of 'SPARQL Query' tab
   INFO  13:06:10  Disposed of workspace
   INFO  13:06:10  
  ERROR  13:06:10  Uncaught Exception in thread 'AWT-EventQueue-0'
java.lang.OutOfMemoryError: Java heap space
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user


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

Re: Loading SNOMED Clinical Findings ontology into Protege

Wynne, Robert (NIH/NLM/LHC) [C]

On Windows, edit your run.bat file to increase your memory settings.

I don’t think any modifications are needed to the .ini (has it been renamed with l4j to show it’s intended for logging only?).

 

Just to add, IHTSDO has a SNOMED to OWL toolkit[1] under continuous development.  The run instructions are very well documented.  SNOMED releases are available from NLM as input[2].

The resulting OWL is EL++ and classifies quickly with the Elk reasoner.

 

The file you downloaded looks like a rough translation from an older Metathesaurus extraction or subset thereof.

If there are areas not covered by the toolkit, another approach you might try would be to run the Metathesaurus extraction through umls2rdf[3].  The resulting format is Turtle RDF.

Turtle can be loaded to triplestores and Protégé.

 

Hope this helps.

Rob

 

[1] https://github.com/IHTSDO/snomed-owl-toolkit

[2] https://www.nlm.nih.gov/healthit/snomedct/index.html

[3] https://github.com/ncbo/umls2rdf

 

From: Samson Tu <[hidden email]>
Sent: Wednesday, November 27, 2019 1:14 PM
To: User support for WebProtege and Protege Desktop <[hidden email]>
Subject: Re: [protege-user] Loading SNOMED Clinical Findings ontology into Protege

 

Hi,

 

I have no problem loading a downloaded copy of SNOMED CF in Protege 5.5 that has the setting of Xmx3000M.

 

SNOMED CF is the extracted “findings” hierarchy of SNOMED CT (as of 2014).

 

I have no problem loading the 2019 version of SNOMED CT with the Xmx5000M setting.

 

I am using a 2013 MacBook Pro.

 

The earlier discussion was about loading SNOMED CT+LOINC+RxNORM.

 

With best regards,

Samson

 



On Nov 27, 2019, at 9:34 AM, Michael DeBellis <[hidden email]> wrote:

 

I realize this is a question more for BioPortal but I recall that some researchers posted to this list a while ago who where were working on SNOMED and last time I asked a question to the Bioportal site I didn't get a very detailed answer (although I'll try them again as well). BTW, if there is a user group or list specifically for SNOMED any links would be appreciated. 

 

I tried loading the OWL file that can be found here: http://bioportal.bioontology.org/ontologies/SNOMED_CF  (the file at the bottom of that page dated 5/7/2014). The file takes a very long time to load but ultimately fails due to an error that I think indicates Java ran out of virtual memory (I'll post the log info below). A while back I was trying to load a large ontology and I needed to increase the parameters on my ini file. My current ini file is:

 

-Xms200M
-Xmx2000M
-Xss16M

 

I'm running Windows 10 Home version on a desktop with 12GB of RAM.

 

My questions are: 

1)  Is it feasible to load this whole ontology on my kind of desktop by further editing my ini file? I could also purchase more RAM if that would help. Any recommendations as to what I should change the ini parameters to to make it work? I have to confess I don't understand what these parameters are doing, I just know someone suggested I increase one of them and I did and it worked on other large ontologies up until now.

 

2) If the ontology is just too large is there a way to extract some parts of it? I would like to do that anyway because I don't really need the whole ontology. I tried using the "File>Open from URL" option and gave it the ID for one high level class from that ontology (the disease class: http://ihtsdo.org/snomedct/clinicalFinding#64572001) I was hoping it might just load that class and all it's subclasses but after churning for a while I got an error that Protege couldn't parse the ontology. 

 

3) Are there other ways to get this to work such as use a triple store? Any suggestions as to tutorials or other info about what or how to use in this case?

 

Here is the log file below. I cleared the warning messages that I get during a normal startup so these are just the messages after trying to load that ontology, this was a fresh start so no issues of other ontologies taking up space. Any feedback would be appreciated:

 

------------------------------ Log file ----------------------------------------------

 

   INFO  12:47:31  OWL API Version: 4.5.9.2019-02-01T07:24:44Z
   INFO  12:47:34  ------------------------------- Loading Ontology -------------------------------
   INFO  12:47:34  Loading ontology from file:/C:/Users/Michael%20DeBellis/Desktop/Ontologies/Neha/snomed_cf.owl/snomed_cf.owl
   INFO  13:06:10  ---------------------------- Disposing of Workspace ----------------------------
   INFO  13:06:10  Finished loading file:/C:/Users/Michael%20DeBellis/Desktop/Ontologies/Neha/snomed_cf.owl/snomed_cf.owl
   INFO  13:06:10  Saved tab state for 'Object properties' tab
   INFO  13:06:10  Saved tab state for 'DL Query' tab
   INFO  13:06:10  Saved tab state for 'Entities' tab
   INFO  13:06:10  Saved tab state for 'Classes' tab
   INFO  13:06:10  Saved tab state for 'SWRLTab' tab
   INFO  13:06:10  Saved tab state for 'Data properties' tab
   INFO  13:06:10  Saved tab state for 'Active ontology' tab
   INFO  13:06:10  Saved tab state for 'SQWRLTab' tab
   INFO  13:06:10  Saved tab state for 'Individuals by class' tab
   INFO  13:06:10  Saved tab state for 'SPARQL Query' tab
   INFO  13:06:10  Saved workspace
   INFO  13:06:10  Disposed of 'Object properties' tab
   INFO  13:06:10  Disposed of 'DL Query' tab
   INFO  13:06:10  Disposed of 'Entities' tab
   INFO  13:06:10  Disposed of 'Classes' tab
   INFO  13:06:10  Disposed of 'SWRLTab' tab
   INFO  13:06:10  Disposed of 'Data properties' tab
   INFO  13:06:10  Disposed of 'Active ontology' tab
   INFO  13:06:10  Disposed of 'SQWRLTab' tab
   INFO  13:06:10  Disposed of 'Individuals by class' tab
   INFO  13:06:10  Disposed of 'SPARQL Query' tab
   INFO  13:06:10  Disposed of workspace
   INFO  13:06:10  
  ERROR  13:06:10  Uncaught Exception in thread 'AWT-EventQueue-0'
java.lang.OutOfMemoryError: Java heap space

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

 


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

Re: Loading SNOMED Clinical Findings ontology into Protege

Michael DeBellis-2
In reply to this post by samsontu
Samson, I had to tweak the variable a couple times but I got it to work.  I think I may have initially over corrected and set it too high. Just FYI, here's what I did:

First, I tried increasing my Xmx setting to  5000M and  that didn't work. Then I went to Help>About (after a clean restart and with the Xmx set to 5000) and the output was:

Version 5.5.0
Max memory set to 444MB (via Java -Xmx setting)
Currently using 40 MB

So I thought maybe I had it set too high for my machine (either due to the amount of RAM or the way my Java VM is configured or ???) and that as a result it was just going for a default. FYI, I'm using Java 8 Update 231 (64 bit) and the Java SE Development kit 8 Update 191 (64 bit). 

So I set it down to 3000M (still more than the 2000M I started with)

and then restarted Protege. When I do that it seemed like Protege was reading the correct number based on the About screen and I was able to load snomed_cf and run the reasoner. It actually loaded fairly quickly. I think you or someone cautioned me about setting the max number too high the last time I had this problem and that's what I originally did but anyway now that it's 3000 rather than 2000 it works. Thanks for the help! Happy Thanksgiving.

Michael



On Wed, Nov 27, 2019 at 10:14 AM Samson Tu <[hidden email]> wrote:
Hi,

I have no problem loading a downloaded copy of SNOMED CF in Protege 5.5 that has the setting of Xmx3000M.

SNOMED CF is the extracted “findings” hierarchy of SNOMED CT (as of 2014).

I have no problem loading the 2019 version of SNOMED CT with the Xmx5000M setting.

I am using a 2013 MacBook Pro.

The earlier discussion was about loading SNOMED CT+LOINC+RxNORM.

With best regards,
Samson


On Nov 27, 2019, at 9:34 AM, Michael DeBellis <[hidden email]> wrote:

I realize this is a question more for BioPortal but I recall that some researchers posted to this list a while ago who where were working on SNOMED and last time I asked a question to the Bioportal site I didn't get a very detailed answer (although I'll try them again as well). BTW, if there is a user group or list specifically for SNOMED any links would be appreciated. 

I tried loading the OWL file that can be found here: http://bioportal.bioontology.org/ontologies/SNOMED_CF  (the file at the bottom of that page dated 5/7/2014). The file takes a very long time to load but ultimately fails due to an error that I think indicates Java ran out of virtual memory (I'll post the log info below). A while back I was trying to load a large ontology and I needed to increase the parameters on my ini file. My current ini file is:

-Xms200M
-Xmx2000M
-Xss16M

I'm running Windows 10 Home version on a desktop with 12GB of RAM.

My questions are: 
1)  Is it feasible to load this whole ontology on my kind of desktop by further editing my ini file? I could also purchase more RAM if that would help. Any recommendations as to what I should change the ini parameters to to make it work? I have to confess I don't understand what these parameters are doing, I just know someone suggested I increase one of them and I did and it worked on other large ontologies up until now.

2) If the ontology is just too large is there a way to extract some parts of it? I would like to do that anyway because I don't really need the whole ontology. I tried using the "File>Open from URL" option and gave it the ID for one high level class from that ontology (the disease class: http://ihtsdo.org/snomedct/clinicalFinding#64572001) I was hoping it might just load that class and all it's subclasses but after churning for a while I got an error that Protege couldn't parse the ontology. 

3) Are there other ways to get this to work such as use a triple store? Any suggestions as to tutorials or other info about what or how to use in this case?

Here is the log file below. I cleared the warning messages that I get during a normal startup so these are just the messages after trying to load that ontology, this was a fresh start so no issues of other ontologies taking up space. Any feedback would be appreciated:

------------------------------ Log file ----------------------------------------------

   INFO  12:47:31  OWL API Version: 4.5.9.2019-02-01T07:24:44Z
   INFO  12:47:34  ------------------------------- Loading Ontology -------------------------------
   INFO  12:47:34  Loading ontology from file:/C:/Users/Michael%20DeBellis/Desktop/Ontologies/Neha/snomed_cf.owl/snomed_cf.owl
   INFO  13:06:10  ---------------------------- Disposing of Workspace ----------------------------
   INFO  13:06:10  Finished loading file:/C:/Users/Michael%20DeBellis/Desktop/Ontologies/Neha/snomed_cf.owl/snomed_cf.owl
   INFO  13:06:10  Saved tab state for 'Object properties' tab
   INFO  13:06:10  Saved tab state for 'DL Query' tab
   INFO  13:06:10  Saved tab state for 'Entities' tab
   INFO  13:06:10  Saved tab state for 'Classes' tab
   INFO  13:06:10  Saved tab state for 'SWRLTab' tab
   INFO  13:06:10  Saved tab state for 'Data properties' tab
   INFO  13:06:10  Saved tab state for 'Active ontology' tab
   INFO  13:06:10  Saved tab state for 'SQWRLTab' tab
   INFO  13:06:10  Saved tab state for 'Individuals by class' tab
   INFO  13:06:10  Saved tab state for 'SPARQL Query' tab
   INFO  13:06:10  Saved workspace
   INFO  13:06:10  Disposed of 'Object properties' tab
   INFO  13:06:10  Disposed of 'DL Query' tab
   INFO  13:06:10  Disposed of 'Entities' tab
   INFO  13:06:10  Disposed of 'Classes' tab
   INFO  13:06:10  Disposed of 'SWRLTab' tab
   INFO  13:06:10  Disposed of 'Data properties' tab
   INFO  13:06:10  Disposed of 'Active ontology' tab
   INFO  13:06:10  Disposed of 'SQWRLTab' tab
   INFO  13:06:10  Disposed of 'Individuals by class' tab
   INFO  13:06:10  Disposed of 'SPARQL Query' tab
   INFO  13:06:10  Disposed of workspace
   INFO  13:06:10  
  ERROR  13:06:10  Uncaught Exception in thread 'AWT-EventQueue-0'
java.lang.OutOfMemoryError: Java heap space
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user

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

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

Re: Loading SNOMED Clinical Findings ontology into Protege

Campbell, James R
Michael
You must have a 64-bit machine running Protege and a version of Java that supports 64 bits to address all that memory with the Java virtual machine.  Otherwise, to run SNOMED CT well I use Xmx=8GB and Xsx=2GB and it works fine. 
Jim Campbell


From: protege-user <[hidden email]> on behalf of Michael DeBellis <[hidden email]>
Sent: Thursday, November 28, 2019 2:38 PM
To: User support for WebProtege and Protege Desktop <[hidden email]>
Subject: Re: [protege-user] Loading SNOMED Clinical Findings ontology into Protege
 
Non-UNMC email
Samson, I had to tweak the variable a couple times but I got it to work.  I think I may have initially over corrected and set it too high. Just FYI, here's what I did:

First, I tried increasing my Xmx setting to  5000M and  that didn't work. Then I went to Help>About (after a clean restart and with the Xmx set to 5000) and the output was:

Version 5.5.0
Max memory set to 444MB (via Java -Xmx setting)
Currently using 40 MB

So I thought maybe I had it set too high for my machine (either due to the amount of RAM or the way my Java VM is configured or ???) and that as a result it was just going for a default. FYI, I'm using Java 8 Update 231 (64 bit) and the Java SE Development kit 8 Update 191 (64 bit). 

So I set it down to 3000M (still more than the 2000M I started with)

and then restarted Protege. When I do that it seemed like Protege was reading the correct number based on the About screen and I was able to load snomed_cf and run the reasoner. It actually loaded fairly quickly. I think you or someone cautioned me about setting the max number too high the last time I had this problem and that's what I originally did but anyway now that it's 3000 rather than 2000 it works. Thanks for the help! Happy Thanksgiving.

Michael



On Wed, Nov 27, 2019 at 10:14 AM Samson Tu <[hidden email]> wrote:
Hi,

I have no problem loading a downloaded copy of SNOMED CF in Protege 5.5 that has the setting of Xmx3000M.

SNOMED CF is the extracted “findings” hierarchy of SNOMED CT (as of 2014).

I have no problem loading the 2019 version of SNOMED CT with the Xmx5000M setting.

I am using a 2013 MacBook Pro.

The earlier discussion was about loading SNOMED CT+LOINC+RxNORM.

With best regards,
Samson


On Nov 27, 2019, at 9:34 AM, Michael DeBellis <[hidden email]> wrote:

I realize this is a question more for BioPortal but I recall that some researchers posted to this list a while ago who where were working on SNOMED and last time I asked a question to the Bioportal site I didn't get a very detailed answer (although I'll try them again as well). BTW, if there is a user group or list specifically for SNOMED any links would be appreciated. 

I tried loading the OWL file that can be found here: http://bioportal.bioontology.org/ontologies/SNOMED_CF  (the file at the bottom of that page dated 5/7/2014). The file takes a very long time to load but ultimately fails due to an error that I think indicates Java ran out of virtual memory (I'll post the log info below). A while back I was trying to load a large ontology and I needed to increase the parameters on my ini file. My current ini file is:

-Xms200M
-Xmx2000M
-Xss16M

I'm running Windows 10 Home version on a desktop with 12GB of RAM.

My questions are: 
1)  Is it feasible to load this whole ontology on my kind of desktop by further editing my ini file? I could also purchase more RAM if that would help. Any recommendations as to what I should change the ini parameters to to make it work? I have to confess I don't understand what these parameters are doing, I just know someone suggested I increase one of them and I did and it worked on other large ontologies up until now.

2) If the ontology is just too large is there a way to extract some parts of it? I would like to do that anyway because I don't really need the whole ontology. I tried using the "File>Open from URL" option and gave it the ID for one high level class from that ontology (the disease class: http://ihtsdo.org/snomedct/clinicalFinding#64572001) I was hoping it might just load that class and all it's subclasses but after churning for a while I got an error that Protege couldn't parse the ontology. 

3) Are there other ways to get this to work such as use a triple store? Any suggestions as to tutorials or other info about what or how to use in this case?

Here is the log file below. I cleared the warning messages that I get during a normal startup so these are just the messages after trying to load that ontology, this was a fresh start so no issues of other ontologies taking up space. Any feedback would be appreciated:

------------------------------ Log file ----------------------------------------------

   INFO  12:47:31  OWL API Version: 4.5.9.2019-02-01T07:24:44Z
   INFO  12:47:34  ------------------------------- Loading Ontology -------------------------------
   INFO  12:47:34  Loading ontology from file:/C:/Users/Michael%20DeBellis/Desktop/Ontologies/Neha/snomed_cf.owl/snomed_cf.owl
   INFO  13:06:10  ---------------------------- Disposing of Workspace ----------------------------
   INFO  13:06:10  Finished loading file:/C:/Users/Michael%20DeBellis/Desktop/Ontologies/Neha/snomed_cf.owl/snomed_cf.owl
   INFO  13:06:10  Saved tab state for 'Object properties' tab
   INFO  13:06:10  Saved tab state for 'DL Query' tab
   INFO  13:06:10  Saved tab state for 'Entities' tab
   INFO  13:06:10  Saved tab state for 'Classes' tab
   INFO  13:06:10  Saved tab state for 'SWRLTab' tab
   INFO  13:06:10  Saved tab state for 'Data properties' tab
   INFO  13:06:10  Saved tab state for 'Active ontology' tab
   INFO  13:06:10  Saved tab state for 'SQWRLTab' tab
   INFO  13:06:10  Saved tab state for 'Individuals by class' tab
   INFO  13:06:10  Saved tab state for 'SPARQL Query' tab
   INFO  13:06:10  Saved workspace
   INFO  13:06:10  Disposed of 'Object properties' tab
   INFO  13:06:10  Disposed of 'DL Query' tab
   INFO  13:06:10  Disposed of 'Entities' tab
   INFO  13:06:10  Disposed of 'Classes' tab
   INFO  13:06:10  Disposed of 'SWRLTab' tab
   INFO  13:06:10  Disposed of 'Data properties' tab
   INFO  13:06:10  Disposed of 'Active ontology' tab
   INFO  13:06:10  Disposed of 'SQWRLTab' tab
   INFO  13:06:10  Disposed of 'Individuals by class' tab
   INFO  13:06:10  Disposed of 'SPARQL Query' tab
   INFO  13:06:10  Disposed of workspace
   INFO  13:06:10  
  ERROR  13:06:10  Uncaught Exception in thread 'AWT-EventQueue-0'
java.lang.OutOfMemoryError: Java heap space
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user

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

The information in this e-mail may be privileged and confidential, intended only for the use of the addressee(s) above. Any unauthorized use or disclosure of this information is prohibited. If you have received this e-mail by mistake, please delete it and immediately contact the sender.
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
Reply | Threaded
Open this post in threaded view
|

Re: Loading SNOMED Clinical Findings ontology into Protege

samsontu
In reply to this post by Michael DeBellis-2
Hi Michael,

On Nov 28, 2019, at 1:38 PM, Michael DeBellis <[hidden email]> wrote:

First, I tried increasing my Xmx setting to  5000M and  that didn't work. Then I went to Help>About (after a clean restart and with the Xmx set to 5000) and the output was:

Version 5.5.0
Max memory set to 444MB (via Java -Xmx setting)
Currently using 40 MB

So I thought maybe I had it set too high for my machine (either due to the amount of RAM or the way my Java VM is configured or ???) and that as a result it was just going for a default. FYI, I'm using Java 8 Update 231 (64 bit) and the Java SE Development kit 8 Update 191 (64 bit). 


That’s very strange. I’ve never seen Protege java process not acknowledging the -Xmx setting when the allocated memory is “too high.” Still, glad that you got it to work.

It would be good, as Rob suggested, to get the more recent version of SNOMED CT than the 2014 (or earlier) extract. We probably should ask Bioportal to take down the SNOMED CF extract. It’s no longer serving the purpose for which we created the extract.

With best regards,
Samson


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

Re: Loading SNOMED Clinical Findings ontology into Protege

Michael DeBellis-2
In reply to this post by Wynne, Robert (NIH/NLM/LHC) [C]
Rob, Thanks very much for that info and those links. I had no idea I was using the wrong version. 

Cheers,
Michael

On Wed, Nov 27, 2019 at 3:29 PM Wynne, Robert (NIH/NLM/LHC) [C] <[hidden email]> wrote:

On Windows, edit your run.bat file to increase your memory settings.

I don’t think any modifications are needed to the .ini (has it been renamed with l4j to show it’s intended for logging only?).

 

Just to add, IHTSDO has a SNOMED to OWL toolkit[1] under continuous development.  The run instructions are very well documented.  SNOMED releases are available from NLM as input[2].

The resulting OWL is EL++ and classifies quickly with the Elk reasoner.

 

The file you downloaded looks like a rough translation from an older Metathesaurus extraction or subset thereof.

If there are areas not covered by the toolkit, another approach you might try would be to run the Metathesaurus extraction through umls2rdf[3].  The resulting format is Turtle RDF.

Turtle can be loaded to triplestores and Protégé.

 

Hope this helps.

Rob

 

[1] https://github.com/IHTSDO/snomed-owl-toolkit

[2] https://www.nlm.nih.gov/healthit/snomedct/index.html

[3] https://github.com/ncbo/umls2rdf

 

From: Samson Tu <[hidden email]>
Sent: Wednesday, November 27, 2019 1:14 PM
To: User support for WebProtege and Protege Desktop <[hidden email]>
Subject: Re: [protege-user] Loading SNOMED Clinical Findings ontology into Protege

 

Hi,

 

I have no problem loading a downloaded copy of SNOMED CF in Protege 5.5 that has the setting of Xmx3000M.

 

SNOMED CF is the extracted “findings” hierarchy of SNOMED CT (as of 2014).

 

I have no problem loading the 2019 version of SNOMED CT with the Xmx5000M setting.

 

I am using a 2013 MacBook Pro.

 

The earlier discussion was about loading SNOMED CT+LOINC+RxNORM.

 

With best regards,

Samson

 



On Nov 27, 2019, at 9:34 AM, Michael DeBellis <[hidden email]> wrote:

 

I realize this is a question more for BioPortal but I recall that some researchers posted to this list a while ago who where were working on SNOMED and last time I asked a question to the Bioportal site I didn't get a very detailed answer (although I'll try them again as well). BTW, if there is a user group or list specifically for SNOMED any links would be appreciated. 

 

I tried loading the OWL file that can be found here: http://bioportal.bioontology.org/ontologies/SNOMED_CF  (the file at the bottom of that page dated 5/7/2014). The file takes a very long time to load but ultimately fails due to an error that I think indicates Java ran out of virtual memory (I'll post the log info below). A while back I was trying to load a large ontology and I needed to increase the parameters on my ini file. My current ini file is:

 

-Xms200M
-Xmx2000M
-Xss16M

 

I'm running Windows 10 Home version on a desktop with 12GB of RAM.

 

My questions are: 

1)  Is it feasible to load this whole ontology on my kind of desktop by further editing my ini file? I could also purchase more RAM if that would help. Any recommendations as to what I should change the ini parameters to to make it work? I have to confess I don't understand what these parameters are doing, I just know someone suggested I increase one of them and I did and it worked on other large ontologies up until now.

 

2) If the ontology is just too large is there a way to extract some parts of it? I would like to do that anyway because I don't really need the whole ontology. I tried using the "File>Open from URL" option and gave it the ID for one high level class from that ontology (the disease class: http://ihtsdo.org/snomedct/clinicalFinding#64572001) I was hoping it might just load that class and all it's subclasses but after churning for a while I got an error that Protege couldn't parse the ontology. 

 

3) Are there other ways to get this to work such as use a triple store? Any suggestions as to tutorials or other info about what or how to use in this case?

 

Here is the log file below. I cleared the warning messages that I get during a normal startup so these are just the messages after trying to load that ontology, this was a fresh start so no issues of other ontologies taking up space. Any feedback would be appreciated:

 

------------------------------ Log file ----------------------------------------------

 

   INFO  12:47:31  OWL API Version: 4.5.9.2019-02-01T07:24:44Z
   INFO  12:47:34  ------------------------------- Loading Ontology -------------------------------
   INFO  12:47:34  Loading ontology from file:/C:/Users/Michael%20DeBellis/Desktop/Ontologies/Neha/snomed_cf.owl/snomed_cf.owl
   INFO  13:06:10  ---------------------------- Disposing of Workspace ----------------------------
   INFO  13:06:10  Finished loading file:/C:/Users/Michael%20DeBellis/Desktop/Ontologies/Neha/snomed_cf.owl/snomed_cf.owl
   INFO  13:06:10  Saved tab state for 'Object properties' tab
   INFO  13:06:10  Saved tab state for 'DL Query' tab
   INFO  13:06:10  Saved tab state for 'Entities' tab
   INFO  13:06:10  Saved tab state for 'Classes' tab
   INFO  13:06:10  Saved tab state for 'SWRLTab' tab
   INFO  13:06:10  Saved tab state for 'Data properties' tab
   INFO  13:06:10  Saved tab state for 'Active ontology' tab
   INFO  13:06:10  Saved tab state for 'SQWRLTab' tab
   INFO  13:06:10  Saved tab state for 'Individuals by class' tab
   INFO  13:06:10  Saved tab state for 'SPARQL Query' tab
   INFO  13:06:10  Saved workspace
   INFO  13:06:10  Disposed of 'Object properties' tab
   INFO  13:06:10  Disposed of 'DL Query' tab
   INFO  13:06:10  Disposed of 'Entities' tab
   INFO  13:06:10  Disposed of 'Classes' tab
   INFO  13:06:10  Disposed of 'SWRLTab' tab
   INFO  13:06:10  Disposed of 'Data properties' tab
   INFO  13:06:10  Disposed of 'Active ontology' tab
   INFO  13:06:10  Disposed of 'SQWRLTab' tab
   INFO  13:06:10  Disposed of 'Individuals by class' tab
   INFO  13:06:10  Disposed of 'SPARQL Query' tab
   INFO  13:06:10  Disposed of workspace
   INFO  13:06:10  
  ERROR  13:06:10  Uncaught Exception in thread 'AWT-EventQueue-0'
java.lang.OutOfMemoryError: Java heap space

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

 

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

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