Stream: genomics/committers
Topic: codings in IG
Bob Milius (Mar 04 2022 at 17:40):
I need to make some corrections in the HLA section in https://build.fhir.org/ig/HL7/genomics-reporting/codings. Should I make the change? or provide the correction so that someone else implements it?
Kevin Power (Mar 04 2022 at 17:43):
@ Bob Milius - Have at it. If it is simple typo type stuff, you can fix it directly in master, or doing a branch is optional if there is a little more to it. If you want to have a tech correction JIRA, that is fine as well.
Kevin Power (Mar 07 2022 at 13:12):
@ Bob Milius - Any update on making this change?
Bob Milius (Mar 07 2022 at 15:57):
I just cloned master and started building it. Build was successful, but noticed errors about RelatedArtifact not handled - should not be here during the build process. Should I ignore those?
e.g.,
Error Generating Narrative for Observation/obs-idh-ex: type org.hl7.fhir.r5.model.RelatedArtifact not handled - should not be here
org.apache.commons.lang3.NotImplementedException: type org.hl7.fhir.r5.model.RelatedArtifact not handled - should not be here
at org.hl7.fhir.r5.renderers.ProfileDrivenRenderer.renderLeaf(ProfileDrivenRenderer.java:416)
at org.hl7.fhir.r5.renderers.ProfileDrivenRenderer.generateElementByProfile(ProfileDrivenRenderer.java:755)
at org.hl7.fhir.r5.renderers.ProfileDrivenRenderer.generateByProfile(ProfileDrivenRenderer.java:678)
at org.hl7.fhir.r5.renderers.ProfileDrivenRenderer.render(ProfileDrivenRenderer.java:120)
at org.hl7.fhir.r5.renderers.ResourceRenderer.render(ResourceRenderer.java:87)
at org.hl7.fhir.igtools.publisher.Publisher.generateNarratives(Publisher.java:1824)
at org.hl7.fhir.igtools.publisher.Publisher.loadConformance(Publisher.java:4671)
at org.hl7.fhir.igtools.publisher.Publisher.createIg(Publisher.java:1005)
at org.hl7.fhir.igtools.publisher.Publisher.execute(Publisher.java:856)
at org.hl7.fhir.igtools.publisher.Publisher.main(Publisher.java:10005)
Kevin Power (Mar 07 2022 at 15:58):
Yup ignore
Bob Milius (Mar 07 2022 at 18:28):
I notice that several of the codesystem URIs on the codings.html page use https instead of http (e.g., https://www.ncbi.nlm.nih.gov/gene/). I believe the terminology group says to use the latter. Should I change them?
Bob Milius (Mar 07 2022 at 18:48):
I pushed my HLA and glstring edits on codings.html to master. I didn't edit any of the other entries.
Kevin Power (Mar 07 2022 at 19:07):
Bob Milius said:
I notice that several of the codesystem URIs on the
codings.htmlpage usehttpsinstead ofhttp(e.g., https://www.ncbi.nlm.nih.gov/gene/). I believe the terminology group says to use the latter. Should I change them?
I say we should change them - @Patrick Werner / @Jamie Jones ?
Kevin Power (Mar 07 2022 at 19:08):
Are you saying our CodeSystem's are defined wrong, or just the codings.html page lists them wrong?
Bob Milius (Mar 07 2022 at 19:46):
Kevin Power said:
Are you saying our CodeSystem's are defined wrong, or just the codings.html page lists them wrong?
Not 'wrong' in the sense of them being resolvable, but I think 'wrong' in the sense of using recommended pattern for coding system URIs by the terminology group. http is usually redirected to https when necessary for those sites.
Bob Milius (Mar 07 2022 at 19:57):
e.g., https://loinc.org will resolve, but won't validate. http://loinc.org will validate and resolve.
Bret H (Mar 07 2022 at 20:02):
is it the codings.html page? or elsewhere?
Bob Milius (Mar 07 2022 at 20:26):
yes, codings.html
Kevin Power (Mar 07 2022 at 21:27):
I think updates to codings.html would be very appropriate.
Patrick Werner (Mar 08 2022 at 11:36):
The canonical urls are defined including the protocol (http/https). Unfortunately.
Patrick Werner (Mar 08 2022 at 11:37):
If we are using Codings which have a defined System URI by HTA and/or FHIR Terminology we have to use this exactly as defined.
Bob Milius (Mar 08 2022 at 15:42):
We have a column in codings.html named "Homepage (FHIR CodeSystem URI)" with entries like "https://www.ncbi.nlm.nih.gov/clinvar (http://www.ncbi.nlm.nih.gov/clinvar)". This is a bit confusing to me. Why not use the http: in the 'Homepage' if it resolves correctly to https? For items that don't have an official FHIR Codesystem URI, like 'https://www.ncbi.nlm.nih.gov/dbvar/ (not assigned yet)", we could use http: if it resolves. (btw, we should remove trailing slashes as well). Maybe bring it up at today's call?
Kevin Power (Mar 08 2022 at 15:50):
I do have it on the list to vote and approve that we submit our publication request, so yes, we probably do need to figure this out.
Kevin Power (Mar 08 2022 at 15:52):
So, the first link should be considered the "click here for the page", and the second link should be the official CS URI. With that, I don't think it would be surprising that they are different. To me, if the website expects HTTPS, we should include HTTPS in the first link.
Bob Milius (Mar 08 2022 at 15:57):
codesystems don't have to resolve, but it sure would be nice if they did.
Kevin Power (Mar 08 2022 at 15:58):
Agreed, and I think that is why we have both - the Home Page link to tell a user where to go to find out more, and the CS URI for official usage.
Maybe we convert the "Homepage" to an actual link?
Kevin Power (Mar 08 2022 at 15:59):
And maybe even move the 'Home Page' URL to the description column, like as a "Click here to learn more"
Bob Milius (Mar 08 2022 at 15:59):
There are bunch of systems that aren't defined by HL7, e.g., many local systems. For those systems in our codings.html that don't have an official HL7 system, what is our recommendation for a codesystem in our IG? to use the home page link?
Kevin Power (Mar 08 2022 at 15:59):
(I do agree that the two things in the same column like that does look a little weird.
Kevin Power (Mar 08 2022 at 16:00):
Bob Milius said:
There are bunch of systems that aren't defined by HL7, e.g., many local systems. For those systems in our
codings.htmlthat don't have an official HL7 system, what is our recommendation for a codesystem in our IG? to use the home page link?
Good question, we probably do not have a recommendation?
Kevin Power (Mar 09 2022 at 17:15):
@ Bob Milius - any idea when you will get these changes drafted and shared for review?
Bob Milius (Mar 09 2022 at 19:26):
Kevin Power said:
Bob Milius - any idea when you will get these changes drafted and shared for review?
I think I should have something ready to review later today or tomorrow morning.
Bob Milius (Mar 10 2022 at 20:19):
fyi, the HTA entry for HGNC is incomplete. It only lists the base url without any differentiation between gene names and gene groups. Our communication with genenames.org led us to recommend using http://www.genenames.org/geneId and http://www.genenames.org/genegroup for these codesystems. I'm not sure how to reconcile this in our IG.
Kevin Power (Mar 10 2022 at 21:15):
@Joel Schneider - what do you think?
Bob Milius (Mar 10 2022 at 21:15):
Basically, I'm including a top level narrative explaining that
Genomics covers a broad landscape involving many organizations outside of HL7 that have established codes and name systems for different use cases. The HL7 Terminology Authority governs the use of terminologies that are external to HL7 and they have begun to document these. The HTA is working with these organizations to establish canonical URIs that can be used as HL7 code and name systems. Many of these are listed in the the core specification guidance.
We have listed many, but not all, of genomics-related systems below. In this table, we provide links for more information -- to the HTA entries when available, or to organization home page if not. We also list a URI that has been approved by the HTA or by Vocabulary work group, and a label defined in the LRI V2 specification.For those organization that list the URI as "not yet assigned," we recommend using the organization site URL, change any "https" to "http", and remove trailing slashes.
Bob Milius (Mar 10 2022 at 21:18):
Also, there's this page from the Vocab WG - https://confluence.hl7.org/display/VOC/External+Code+System+URI+Quality+Assurance
which includes many (perhaps all) of the codesystems listed in the https://hl7.org/fhir/terminologies-systems.html. If it's in one of those, I'll use it. I posted a comment re HTA HGNC page re gene names and gene groups. I say we go with what we determined from our communications with genenames.org, and let HTA catch up. They have a lot of work to do and we should help out where we can.
Bob Milius (Mar 10 2022 at 21:21):
Sorry, this is taking a bit longer than I expected, but should have something to share end of today.
Bob Milius (Mar 10 2022 at 22:23):
I created a Jira ticket (FHIR-36251]), created a branch called FHIR-36251-externalcodesystems and pushed my changes to codings.html to it. I think we need to resize the columns. And perhaps make the home page links clickable. Should I do a pull request?
Joel Schneider (Mar 10 2022 at 22:24):
Bob Milius said:
fyi, the HTA entry for HGNC is incomplete.
Looks like the HGNC page does not specify a canonical url, and also is not using the latest template (e.g. comparing it to the HTA entry for ClinVar Variation ID).
We may want to open a HTA ticket to address the HGNC issues.
Bob Milius (Mar 10 2022 at 22:27):
Agreed. When HTA doesn't have a URI, I look at both
- https://confluence.hl7.org/display/VOC/External+Code+System+URI+Quality+Assurance
- https://build.fhir.org/terminologies-systems.html
If it's there's a URI in one of those, I use that in the codings.html page.
Joel Schneider (Mar 11 2022 at 15:35):
Open ticket for the HGNC issue (created by Patrick in January): HTA-71
Bob Milius (Mar 11 2022 at 18:26):
submitted pull request for FHIR-36251-externalcodesystems branch
Kevin Power (Mar 11 2022 at 19:22):
I did a quick review, and it looks good to me - thanks @ Bob Milius !
I tagged @Jamie Jones, @Patrick Werner, and @Bret H on the PR. I don't see @Joel Schneider on github, but it would be great for you to review as well.
Joel Schneider (Mar 12 2022 at 04:03):
HL7's intention is to eventually have terminology.hl7.org (a.k.a. THO) become the authoritative source for code system information.
The THO External Terminologies page already lists several genomics-related code systems, such as GTR, Hgnc, Hgvs, and ISCN.
https://terminology.hl7.org/external_terminologies.html
It would be best to prominently mention the terminology.hl7.org site on the CG IG's codings page.
Joel Schneider (Mar 12 2022 at 04:30):
I also left a comment on the PR.
Patrick Werner (Mar 14 2022 at 09:44):
Added a comment about Gene Code
(NCBI/NLM). We don't use this in the IG, so we could just remove it to avoid confusion with HGNC Gene ID
Bob Milius (Mar 14 2022 at 14:05):
thanks @Joel Schneider and @Patrick Werner . I'll make the changes today.
Bob Milius (Mar 14 2022 at 17:40):
I just pushed changes to the the branch. Please take a look.
Kevin Power (Mar 14 2022 at 18:25):
One minor suggestion (use confluence link instead of older HL7.org link)
Bob Milius (Mar 14 2022 at 21:05):
Kevin Power said:
One minor suggestion (use confluence link instead of older HL7.org link)
I thought about that, but the confluence link doesn't point to any co-chair emails, while the CG leadership page on hl7.org site does.
Bob Milius (Mar 15 2022 at 13:57):
I changed the link to point to the CG confluence page.
Kevin Power (Mar 15 2022 at 13:57):
I just made some updates to the CG confluence page as well.
Bob Milius (Mar 15 2022 at 13:57):
@Bret H do you plan on reviewing the branch, or can I merge with master now?
Bob Milius (Mar 15 2022 at 14:32):
going to merge with master so we have something to share at today's call
Bob Milius (Mar 15 2022 at 14:37):
curious... how long does it take for a build after merge? I assume it happens automagically
Bob Milius (Mar 15 2022 at 14:38):
see it now! http://build.fhir.org/ig/HL7/genomics-reporting/codings.html
Kevin Power (Mar 15 2022 at 14:38):
It is auto, but usually 1-5 minutes
Bob Milius (Mar 15 2022 at 16:03):
I plan on editing master directly to make the HGNC text change we voted on today.
Bob Milius (Mar 15 2022 at 17:11):
pushed to master!
Bob Milius (Mar 15 2022 at 18:00):
Sorry, I'd like to make one more minor edit. I'd like to change "Older systems might send HGNC codes without the 'HGNC: prefix" to "Older systems might send HGNC gene IDs without the 'HGNC: prefix", for clarification.
Bob Milius (Mar 15 2022 at 19:58):
pushed to master!
Bret H (Mar 21 2022 at 12:34):
I see an entry for 'Gene Code (NCBI/NLM)' on http://build.fhir.org/ig/HL7/genomics-reporting/codings.html should it be there?
Kevin Power (Mar 21 2022 at 14:28):
@ Bob Milius - It seems we talked about that one, but I can't remember. I think I recall we were going to remove it, but am not sure.
Patrick Werner (Mar 23 2022 at 09:46):
+1 for removing it, as we are not using it in the IG
Bob Milius (Mar 23 2022 at 12:29):
Bret H said:
I see an entry for 'Gene Code (NCBI/NLM)' on http://build.fhir.org/ig/HL7/genomics-reporting/codings.html should it be there?
We have text somewhere saying that if an HGNC code isn't available, NCBI/NLM may have one to use. Patrick suggested removing it but after finding the bit of narrative he agreed that we leave it in during one of our calls.
just found it here for for value[x] in component:gene-studied
in https://build.fhir.org/ig/HL7/genomics-reporting/StructureDefinition-finding.html
The HGNC gene symbol is to be used as display text and the HGNC gene ID used as the code. If no HGNC code issued for this gene yet, NCBI gene IDs SHALL be used.
Last updated: Apr 12 2022 at 19:14 UTC
