search

NIH-NCPI / ncpi-fhir-ig

🔥 NCPI FHIR Implementation Guide
https://nih-ncpi.github.io/ncpi-fhir-ig/
Creative Commons Zero v1.0 Universal
5 stars 2 forks source link

ResearchStudy.condition coding inconsistencies between dbGap, AnVIL and KidsFirst #44

Open bwalsh opened 2 years ago

bwalsh commented 2 years ago

We should use the same ontologies (or provide mappings). While all three systems use ResearchStudy.condition to indicate the condition that is the focus of the study. All three have a significant number of studies without a populated condition. In addition, all three systems use different ontologies:

system: ontology system: study_count

* 'kidsfirst': 
    * None: 21,
    * 'http://snomed.info/sct': 8
* 'dbgap':
    * None: 234,
    * 'https://dbgap-api.ncbi.nlm.nih.gov/fhir/x1/NamingSystem/MeshEntryTerm': 1597,
    * 'https://uts.nlm.nih.gov/metathesaurus.html': 1459,
    * 'urn:oid:2.16.840.1.113883.6.177': 1591    
* 'anvil':
    * None: 358,
    * 'http://purl.obolibrary.org/obo/doid.owl': 35
linikujp commented 2 years ago

Hi Brian, the FHIR ResearchStudy condition and focus are two different things. Some studies do not have a condition, because the study does not focus on conditions, rather than normal sample as a control or reference. We have explored this when looking at the NCPI data catalog.

Thanks, Asiyah

bwalsh commented 2 years ago

Hi Brian, the FHIR ResearchStudy condition and focus are two different things. Some studies do not have a condition, because the study does not focus on conditions, rather than normal sample as a control or reference. We have explored this when looking at the NCPI data catalog.

Thanks, Asiyah

@linikujp

Hi Asiyah,

Agreed that condition and focus are two different things - that may explain at least some of the studies without a Condition.

However, where there is a condition, I'm concerned that we have 5 different ontology systems.

linikujp commented 2 years ago

Hi Brian, Can you list the 5 different ontology systems? I have done some work of mapping DO-MONDO-MeSH for NCPI data catalog. which is based on the dbGaP FHIR server. I also have some experience with ontology harmonization or alignment, and I am working with DO and MONDO to resolve the inconsistencies I see from NCPI use case. If you list the 5 ontologies out, we may make a case for aligning all these 5 ontologies for the NCPI community. Thanks, Asiyah

bwalsh commented 2 years ago

If you list the 5 ontologies out, we may make a case for aligning all these 5 ontologies for the NCPI community.

From above ... Thanks! -b

linikujp commented 2 years ago

Do you know what this is for: urn:oid:2.16.840.1.113883.6.177

On Thu, Dec 2, 2021, 18:31 Brian @.***> wrote:

If you list the 5 ontologies out, we may make a case for aligning all these 5 ontologies for the NCPI community.

From above ... Thanks! -b

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/NIH-NCPI/ncpi-fhir-ig/issues/44#issuecomment-985085880, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACCPEPIDBFY5R4BEOXZM2DLUO76UFANCNFSM5JFVHB6A . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

bwalsh commented 2 years ago

I'm traveling now, and will be out of office next week. Will follow up then. As far as I know:

https://www.hl7.org/documentcenter/public/standards/vocabulary/vocabulary_tables/infrastructure/vocabulary/voc_ExternalSystems.html

On Thu, Dec 2, 2021, 4:21 PM Asiyah Yu Lin @.***> wrote:

Do you know what this is for: urn:oid:2.16.840.1.113883.6.177

On Thu, Dec 2, 2021, 18:31 Brian @.***> wrote:

If you list the 5 ontologies out, we may make a case for aligning all these 5 ontologies for the NCPI community.

From above ... Thanks! -b

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub < https://github.com/NIH-NCPI/ncpi-fhir-ig/issues/44#issuecomment-985085880 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/ACCPEPIDBFY5R4BEOXZM2DLUO76UFANCNFSM5JFVHB6A

. Triage notifications on the go with GitHub Mobile for iOS < https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675

or Android < https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub .

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/NIH-NCPI/ncpi-fhir-ig/issues/44#issuecomment-985108170, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAALVQGSNB2ZVQYU4L5CPBTUPAEO5ANCNFSM5JFVHB6A . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

RobertJCarroll commented 2 years ago

It looks like that urn is mesh again.

Based on our prior conversations, it sounded like "Given we are using these ResearchStudy objects as a way to do information retrieval, MeSH is a good target". The field is a codeable concept, so we can have a custom free text and multiple ancillary concepts included. IE, if I'd like to flag with SNOMED concept, great, but I should also include the MeSH term that's most relevant.