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] Use of SNOMED CT in virtual appliance

Melissa M mmorine at gmail.com
Fri Feb 8 08:04:39 PST 2019


Hi Clement,
Thanks for your reply. Here are a few examples:

Looking at the FoodOn entry here
<http://bioportal.bioontology.org/ontologies/FOODON?p=classes&conceptid=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FFOODON_03414351>
(which
comes from the langual_import.owl file), I find that searching the
"Preferred Name" and "Synonyms" (which has the same entries as "has narrow
synonym") fields in the annotator do produce results. However none of the
synonyms in the "has synonym" field produce any result.

An example from the base foodon.owl file is here
<http://bioportal.bioontology.org/ontologies/FOODON?p=classes&conceptid=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FFOODON_03420109>.
The preferred name is searchable but "has synonym" field is not.

I found that the entries that are pulled in from the ncbi_taxon.owl import
do not have the same problem. For example in this
<http://bioportal.bioontology.org/ontologies/FOODON?p=classes&conceptid=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FNCBITaxon_3999>
entry,
both the preferred name and synonyms are searchable, however the synonyms
are tagged as "hasExactSynonym" instead.

So basically the problem seems to be related to the "has synonym" field,
and exists across multiple files (that is, both the base foodon.owl and at
least langual_import.owl). I've looked in the .owl files to see if there
are obvious differences in the metadata for these different fields and
don't see anything, but then I'm still learning and am not sure exactly
what I'm looking for.

Thank you for your help and have a nice weekend.
Melissa


On Fri, 8 Feb 2019 at 05:41, Clement Jonquet <jonquet at lirmm.fr> wrote:

> Hi Melissa, all,
>
> See inline.
>
> Le 7 févr. 2019 à 12:22, Melissa M <mmorine at gmail.com> a écrit :
>
> Hi Clement,
> Thanks very much for your reply.
>
> Regarding the FoodOn issue, I should mention 2 things: 1) I have the same
> problem both on my VA and on bioportal, and 2) synonyms are properly
> recognized for some classes and not for others. I've looked at the .owl
> files to see if I could find any differences in specification of classes
> whose synonyms are recognized and those whose synonyms aren't recognized,
> and I couldn't find anything. I'd be happy to provide a few class names as
> examples if you have time to have a look?
>
>
> Please do so.
>
> About the SNOMED issue, I contacted IHTSDO and just heard back this
> morning that the "toolkit is only currently capable of including the Fully
> Specified Name of concepts" but they're opening my issue as a new feature.
> In the meantime I'll start working through the pipeline that you use for
> importing UMLS.
>
>
> Interesting ;)
> It means it worth asking. Indeed, there is n o point doing a semantic
> resource, if it does not include the synonyms.
>
>
> Alternatively, would you be able to provide access to the SNOMED .ttl or
> are there license restrictions?
>
> Thanks again for your help!
> Melissa
>
>
>
> On Thu, 7 Feb 2019 at 03:24, Clement Jonquet <jonquet at lirmm.fr> wrote:
>
>> Melissa,
>>
>> A complementary note:
>>
>> @misha, @john
>> BioPortal takes SNOMED-CT from UMLS with an export from the UMLS2RDF tool
>> as documented.
>> But eventually, now that Snomed International (new name for IHTSDO)
>> provide an OWL file directly, it will be interesting one day to drop the
>> import from UMLS and take the OWL directly.
>> NCBO already does it for NCIt, GO, and a bunch of others that are in UMLS
>> but not imported thru the UMLS export.
>> MeSH is also concerned as there is now an OWL file produced by NLM.
>>
>> @melissa
>> I bet the problem you have with the synonyms in the OWL version of
>> SNOMED-CT is the same that the one for FoodOn. Be sure to specify the right
>> property for synonyms in the « Edit submission information » page.
>>
>> Bye
>> Clement
>>
>> -------------------------------------------------------------------------------------------
>> Dr. Clement JONQUET  -  PhD in Informatics  -  Assistant Professor
>> University of Montpellier (LIRMM)
>>
>> -------------------------------------------------------------------------------------------
>>
>> Le 7 févr. 2019 à 02:02, Michael Dorf <mdorf at stanford.edu> a écrit :
>>
>> Hi Melissa,
>>
>> Yes, that is the script that is being used to import UMLS ontologies into
>> a virtual appliance.  It hasn’t been updated since 2016, but it should
>> still be valid. Let us know if you run into any issues importing.
>>
>> Thanks,
>>
>> Michael
>>
>>
>>
>> On Feb 6, 2019, at 12:10 PM, Melissa M <mmorine at gmail.com> wrote:
>>
>> Hi Michael,
>> Thank you very much for your response. Can you confirm that the pipeline
>> described here
>> <https://www.bioontology.org/wiki/Importing_UMLS_To_Virtual_Appliance> is
>> the one currently used to import UMLS content into bioportal? I'm assuming
>> that NCBO does not provide downloads of the output files from this pipeline
>> (I think in ttl format)?
>> Thanks again,
>> Melissa
>>
>>
>> On Wed, 6 Feb 2019 at 15:42, Michael Dorf <mdorf at stanford.edu> wrote:
>>
>>> Hi Melissa,
>>>
>>> Thanks for contacting us again. Twice per year we do a bulk import of
>>> terminologies distributed by the UMLS [1].  (They have a twice per year
>>> release schedule, e.g., for this year it will be 2019AA and 2019AB).
>>> SNOMEDCT is one of the terminologies contained in the UMLS distribution.
>>> So, the version numbers for SNOMEDCT that you see in BioPortal correspond
>>> to UMLS releases, e.g., 2018AA, 2018AB, etc. More info on their
>>> distributions can be found here:
>>>
>>> https://www.nlm.nih.gov/research/umls/sourcereleasedocs/index.html
>>>
>>> The page contains an alphabetical list of all the terminologies
>>> contained in the UMLS.
>>>
>>> Unfortunately, I’m not familiar with the IHTSDO toolkit, but I can only
>>> assume that may contain either a different version from the one we have in
>>> BioPortal; or a version that may not have all the properties fully
>>> populated.
>>>
>>> Thanks,
>>>
>>> Michael
>>>
>>> [1] https://www.nlm.nih.gov/research/umls/
>>>
>>>
>>>
>>>
>>> On Feb 6, 2019, at 8:28 AM, Melissa M <mmorine at gmail.com> wrote:
>>>
>>> Hello,
>>> I've recently been discussing with your team regarding use of FoodOn in
>>> the virtual appliance but thought I'd write separately about this question
>>> since it's unrelated to the other.
>>> So far I've been able to import ontologies and perform annotations more
>>> or less successfully, apart from the FoodOn issue that I've already written
>>> about. Now I'm trying to import SNOMED CT but am noticing some issues.
>>> First, I converted the SNOMED RF2 snapshot files to owl using the SNOMED
>>> OWL toolkit jar provided by IHTSDO
>>> <https://github.com/IHTSDO/snomed-owl-toolkit>, then imported the
>>> resulting owl into my virtual appliance using default settings (i.e.,
>>> selected OWL format, and no changes to prefLabel/altLabel/etc properties).
>>> The resulting ontology appears to contain all classes in the class browser,
>>> and I can annotate text that contains exact prefLabels of the SNOMED
>>> ontology, but it's missing many of the fields that are present in the
>>> SNOMED instance on the bioportal server (such as synonyms, altLabels,
>>> etc.). I therefore can't annotate any text that contains SNOMED altLabels
>>> and synonyms. I've attached a screenshot of the SNOMED class browser in my
>>> virtual appliance to hopefully show what I mean.
>>> So I'm trying to figure out if I went wrong in my conversion to owl, or
>>> import into the VA.
>>> My question is, was the instance of SNOMED on the bioportal server
>>> produced using the toolkit from IHTSDO? And when importing into bioportal,
>>> were non-default settings used?
>>> Thanks in advance for any info,
>>> Melissa
>>>
>>> <Screen Shot 2019-02-06 at 11.54.52 AM.png>
>>> _______________________________________________
>>> bioontology-support mailing list
>>> bioontology-support at lists.stanford.edu
>>> https://mailman.stanford.edu/mailman/listinfo/bioontology-support
>>>
>>>
>>>
>> _______________________________________________
>> bioontology-support mailing list
>> 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/20190208/3e0d1b39/attachment.html>


More information about the bioontology-support mailing list