retired concepts #21
Comments
|
@zzgulu: could you give a more specific description with regard to the SNOMED Browser? What is an example of a term name we could look at, and where in their UI are they displaying the suggested code replacements? Is it on one of the tabs for a term, e.g., Summary or Details? |
|
Hi Jennifer,
Thank you very much for looking into this. The example that I was looking at is the SNOMED concept 119424003 (Dermoid cyst of ovary) which seems it became retired/inactive in 2016AB. UMLS mrconso table (2017AA) also shows this concept is retired/suppressed.Looking up this concept in IHTSDO brower 2017AA/AB version confirms the retired status and provide more information under Concept details/Refsets/Associated Refset Name section as:
Possibly Equivalent To association reference set (foundation metadata concept) = Benign Teratoma of Ovary (disorder). (please see the screenshot attached)
I have seen similar information for the retired concepts in commercial SNOMED browsers such as Health Language and CliniClue as well.
Thanks again for looking into this
On Thursday, September 7, 2017 1:48 PM, Jennifer Vendetti <[email protected]> wrote:
@zzgulu: could you give a more specific description with regard to the SNOMED Browser? What is an example of a term name we could look at, and where in their UI are they displaying the suggested code replacements? Is it on one of the tabs for a term, e.g., Summary or Details?—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or mute the thread.
|
|
OK - I was able to see this concept in the SNOMED Browser after finding the Option drop down box that allows you to search both active and inactive concepts. Your screen shot attachment didn't come through, but I'm including one here that represents what I'm seeing.
|
|
Yes, that's the property I was looking at.
And my question in the bioportal forum is somehow related. If I look for this concept (119424003) under the regular search UI with all property values checked, nothing comes up. Using the API (if I am using it correctly here) also doesn't show any results ,which made me think perhaps all concepts included in the SNOMED.ttl of Bioportal are active concepts (part of the python code where it declares suppress=N).
http://data.bioontology.org/ontologies/SNOMEDCT/classes/http%3A%2F%2Fpurl.bioontology.org%2Fontology%2FSNOMEDCT%2F119424003?include=all&apikey=0dd892d5-8150-4b13-bc45-8cb246b74744
Another observation that I made is that I can see this code in the bioportal sparql endpoint as being retired (concept status=0) and somehow see some relevant properties (to SNOMED browser) such as "inverse_maybe_a". Is the sparql end point use a different version/format of SNOMED on bioportal?
On Thursday, September 7, 2017 4:17 PM, Jennifer Vendetti <[email protected]> wrote:
OK - I was able to see this concept in the SNOMED Browser after finding the Option drop down box that allows you to search both active and inactive concepts. Your screen shot attachment didn't come through, but I'm including one here that represents what I'm seeing.—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or mute the thread.
|
|
BioPortal only serves the concepts that are contained in the most recent submission of any given ontology. The TTL files generated by this project only contain active concepts. The last time we performed an import of UMLS data was from the 2016AB release, so whatever concepts are active in that release are what BioPortal serves (both in the UI and the REST API). I wasn't on the team during the original BioPortal design phase, but my understanding is that serving content from all submissions of all ontologies in the application wouldn't scale well. We currently have close to 900 ontologies in BioPortal, some of which have hundreds of individual submissions. @msalvadores - feel free to correct me if you think I misrepresented BioPortal design in this paragraph. The BioPortal SPARQL endpoint was published several years ago as a beta service, and subsequently abandoned because we simply don't have the resources to maintain. We have an IT person that restarts the system if it happens to go down, but the actual ontology content hasn't been updated in roughly 2.5 years (perhaps longer). |
|
Thank you very again Jennifer for the explanation. This explain why I wasn't able to bring that concept in API. I generated the owl version from the raw SNOMED files (UMLS direct download and from the perl script included in the distribution) and it was loaded in the Bioportal. However, that version didn't include the retired concepts either since it work on the snapshot subset.
I guess for our use case at Vanderbilt, we can assume whatever concept code that is not returned from the API call is retired ( for a given SNOMED version). We can manually examine those codes external to Bioportal for possible replacements.
Thanks again ! This was very helpful.
On Thursday, September 7, 2017 5:05 PM, Jennifer Vendetti <[email protected]> wrote:
BioPortal only serves the concepts that are contained in the most recent submission of any given ontology.The TTL files generated by this project only contain active concepts. The last time we performed an import of UMLS data was from the 2016AB release, so whatever concepts are active in that release are what BioPortal serves (both in the UI and the REST API).I wasn't on the team during the original BioPortal design phase, but my understanding is that serving content from all submissions of all ontologies in the application wouldn't scale well. We currently have close to 900 ontologies in BioPortal, some of which have hundreds of individual submissions. @msalvadores - feel free to correct me if you think I misrepresented BioPortal design in this paragraph.The BioPortal SPARQL endpoint was published several years ago as a beta service, and subsequently abandoned because we simply don't have the resources to maintain. We have an IT person that restarts the system if it happens to go down, but the actual ontology content hasn't been updated in roughly 2.5 years (perhaps longer).—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or mute the thread.
|
|
one last thought,
Do you think if I just add "suppress = 'O' OR anywhere in the code where it says suppress ='N' in the umls2rdf code I will get the retired concepts in the output? I am not sure if it would be as simple as I am suggesting though. Our use case is only for a couple of terminologies like SNOMED, LOINC, and RxNorm
On Thursday, September 7, 2017 5:22 PM, Sina Madani <[email protected]> wrote:
Thank you very again Jennifer for the explanation. This explain why I wasn't able to bring that concept in API. I generated the owl version from the raw SNOMED files (UMLS direct download and from the perl script included in the distribution) and it was loaded in the Bioportal. However, that version didn't include the retired concepts either since it work on the snapshot subset.
I guess for our use case at Vanderbilt, we can assume whatever concept code that is not returned from the API call is retired ( for a given SNOMED version). We can manually examine those codes external to Bioportal for possible replacements.
Thanks again ! This was very helpful.
On Thursday, September 7, 2017 5:05 PM, Jennifer Vendetti <[email protected]> wrote:
BioPortal only serves the concepts that are contained in the most recent submission of any given ontology.The TTL files generated by this project only contain active concepts. The last time we performed an import of UMLS data was from the 2016AB release, so whatever concepts are active in that release are what BioPortal serves (both in the UI and the REST API).I wasn't on the team during the original BioPortal design phase, but my understanding is that serving content from all submissions of all ontologies in the application wouldn't scale well. We currently have close to 900 ontologies in BioPortal, some of which have hundreds of individual submissions. @msalvadores - feel free to correct me if you think I misrepresented BioPortal design in this paragraph.The BioPortal SPARQL endpoint was published several years ago as a beta service, and subsequently abandoned because we simply don't have the resources to maintain. We have an IT person that restarts the system if it happens to go down, but the actual ontology content hasn't been updated in roughly 2.5 years (perhaps longer).—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or mute the thread.
|
|
Sincere apologies that I don't know the answer off the top of my head. The developer of the code in this particular repository is no longer working on BioPortal. You could certainly fork the code and try making that modification to see what you get. If the resulting TTL does end up containing the retired concepts, it may load into BioPortal without issue. |
If remove all the instances of suppress then you will get all concepts (both current and obsolete). If you keep |
|
Thanks !
I'll try to include obsolete concepts and report back.
So far I am having difficulty loading the regular SNOMED ttl into our local instance, perhaps due do to its size (1.2GB). Compressing and submitting it into the Bioportal didn't help either. Looking at the log files ...
… On Sep 7, 2017, at 9:02 PM, Manuel Salvadores ***@***.***> wrote:
Do you think if I just add "suppress = 'O' OR anywhere in the code where it says suppress ='N' ...
If remove all the instances of suppress then you will get all concepts (both current and obsolete). If you keep suppress = 'O' then you will only get obsolete.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or mute the thread.
|
Would it possible to add an option to the config file for including retired concepts/codes in the output? This is specially useful for SNOMED. The IHTSDO SNOMED browser not only shows a given code has been retired but also the suggested replacement(s) through the ref set properties. Thanks!
The text was updated successfully, but these errors were encountered: