FHIR Chat · Implication v2 · genomics/committers

Stream: genomics/committers

Topic: Implication v2

Kevin Power (Nov 23 2021 at 16:37):

Starting a new thread here to discuss the changes to make for all the implication updates. My suggestion:

• I will merge the File Attachment branch into master before we start making the implication changes.
• I would love to merge the withoutGrouper branch as well, but given our current course and speed, I am not sure I want to wait that long.
• Same for Variant updates, but those will be localized enough that it won't like be a problem.
• I will plan to create the implication-v2 branch sometime soon, and share a link here.

Bret H (Nov 23 2021 at 17:05):

for my lazy fingers, could you put the branch link in this chat?

Kevin Power (Nov 23 2021 at 17:07):

I will when I create it (which I haven't yet)

Kevin Power (Nov 23 2021 at 17:35):

I have setup this JIRA query which should show the implication related items that require a change that we voted on today:
https://jira.hl7.org/issues/?filter=16204
You will have to be logged into JIRA in order to view the items.

Kevin Power (Nov 23 2021 at 18:47):

Branch is here:
https://github.com/HL7/genomics-reporting/tree/implications-updates

I haven't made any changes from master yet.

Kevin Power (Nov 23 2021 at 18:48):

If anyone wants to work on implementing these changes, please assign yourself to one of the JIRAs before making the changes. Chat here with any questions.

Jamie Jones (Nov 23 2021 at 21:03):

10/10 thread so far, thanks Kevin! Got time to grab a couple low-hanging fruit here, unless it would be better for me to work on grouperless so we can get that reviewed further

Kevin Power (Nov 23 2021 at 21:18):

I propose we try to clean up groupless and even variant guidance first, then move onto implication updates.

Kevin Power (Nov 23 2021 at 21:22):

But, it all has to get done :smile:

Jamie Jones (Nov 23 2021 at 21:25):

digging in to 3 easier ones here while I check if Bret has anything to push on grouperless

Kevin Power (Nov 23 2021 at 21:27):

I did make several comments on the variant guidance google doc btw :smile:

Jamie Jones (Nov 23 2021 at 21:47):

testing a build with the added codesystem and renamed components, assigned those 3 trackers to myself

Bret H (Nov 23 2021 at 21:52):

I took an easier one and a harder one. The new extension and providing an example of polygenic risk score. Not sure which is the easier one :grinning: I will be able to address them through tomorrow and will post Thursday. Is that okay time wise or are u hoping for faster?

Jamie Jones (Nov 23 2021 at 22:08):

@Bret H i think it's fine if those are a little delayed. Agree with Kevin I'd rather have the grouperless and variant branches cleaner first so we can vote on them

Jamie Jones (Nov 23 2021 at 22:51):

pushed changes for FHIR#16082, FHIR#32709, FHIR#32794. Will need to update the EA UML on tx-Imp-intro

Kevin Power (Nov 23 2021 at 22:54):

Update the EA UML or redo in Lucid?

Jamie Jones (Nov 23 2021 at 22:54):

Happy to draft it in lucid

Kevin Power (Nov 23 2021 at 22:56):

That is my preference, if we can pull it off.

Bret H (Dec 08 2021 at 16:52):

live but nothing pushed yet.

Kevin Power (Dec 08 2021 at 17:04):

What do you mean @Bret H ?

Bret H (Dec 08 2021 at 17:07):

oh, just letting you know I'm still intending to add to the branch. I mean the effort is live. apologies.

Kevin Power (Dec 08 2021 at 22:21):

Just checking on this one @Bret H and @Jamie Jones - Are the assignees for the JIRAs listed on this report still accurate and reflect the ones you are working on?
https://jira.hl7.org/issues/?filter=16204

Bret H (Dec 08 2021 at 22:44):

There's now only one without an assignee. I think it fits in with the 'predicted-therapeutic-implication' component work

Jamie Jones (Dec 09 2021 at 17:04):

Most of the ones I assigned myself were "done" in that push but I didn't want to mark them as applied until the branch was merged with build

Jamie Jones (Dec 09 2021 at 17:05):

Is it still a branch or is it live now?

Kevin Power (Dec 09 2021 at 17:06):

Still a branch, haven't merged to master yet.

Bret H (Dec 10 2021 at 18:13):

will push something today. I am relooking at the work I did on examples. but could push the valueset, extension, guidance changes now. I am just not happy with the examples - I want to find as simplistic and non-controversial approach as possible.

Jamie Jones (Dec 10 2021 at 18:42):

I drafted the value set on the last push there - FHIR#32709, FHIR#16082, and FHIR#32794 should be good to go, structurally

Kevin Power (Dec 10 2021 at 18:51):

Yup, I see the code system here:
http://build.fhir.org/ig/HL7/genomics-reporting/branches/implications-updates/CodeSystem-genetic-therapeutic-implications-cs.html

And value set:
http://build.fhir.org/ig/HL7/genomics-reporting/branches/implications-updates/ValueSet-genetic-therapeutic-implications-vs.html

Bret H (Dec 10 2021 at 23:27):

I have not pushed the value set FHIR-33009 yet. It will be in the push for FHIR-33009, FHIR-34277, FHIR-28404, FHIR-26426, FHIR-32093

Jamie Jones (Dec 10 2021 at 23:28):

Great! Happy to review

Bret H (Dec 10 2021 at 23:29):

I am pretty sure FHIR-19844 is covered by existing examples. Right now no one is assigned it.

Kevin Power (Dec 10 2021 at 23:37):

@Bret H - I am sure you did, but make sure you sync up with the branch before pushing.

Bret H (Dec 13 2021 at 13:20):

ok.

Bret H (Dec 13 2021 at 21:57):

FYI I'm meeting Bob D tomorrow, will review some of the implication branch with him.

Kevin Power (Dec 13 2021 at 22:06):

Is everything completed? @Jamie Jones ?

Jamie Jones (Dec 13 2021 at 22:10):

Was waiting to review a @Bret H push before committing more

Kevin Power (Dec 13 2021 at 22:10):

I see a commit on GitHub from Jamie from a while back, but nothing from you Bret? Would you just show everything locally? I would actually prefer to see a published build, if that isn't asking too much?

Bret H (Dec 14 2021 at 11:45):

I am reviewing local changes before pushing with him. @Jamie Jones I thought you were finished? are you pushing figures? go ahead and make your commits. It'll be easier to reconcile the changes. I won't be meeting with him till late in the morning.

Kevin Power (Dec 14 2021 at 14:20):

Well, FWIW, I like the "push early and often" approach. The longer you work on something without pushing, the higher likelihood that something will conflict, especially when multiple people are contributing to the same branch.

Bret H (Dec 14 2021 at 19:43):

Spoke with @Bob Dolin he gave a thumbs up and agreed to review before merge with Main. I was mostly concerned about the Risk Assessment used for the Polygenic Risk Score example. I will push what I have currently. I'm still cooking examples for clinical trials, but shouldn't take long.

Bret H (Dec 14 2021 at 20:31):

@Jamie Jones I pushed intermediate changes, be sure to fetch them.

Kevin Power (Dec 15 2021 at 18:03):

Just a couple of quick notes @Bret H
1 - Looks like you tried to add some structure for the 'AND/OR' section, but unless you do additional <p> ... </p> your structure in the XML files won't carry over to the HTML.
2 - So we don't duplicate that text, should we put that text in the 'genomics implication' profile, then perhaps have a simple pointer from the diagnostic and therapeutic profiles?

Bret H (Dec 15 2021 at 19:18):

@Kevin Power I was forced to push without it being done-done. :grinning: tried to convey that in the GIT comments. I am aware that there is formatting to finish

Bret H (Dec 15 2021 at 19:21):

For 2, How about all three. I don't mind the duplication of text and it is an important topic. But did you catch that it is in the general page too (as we spoke about)? What do you think? Is it going to be obvious to implementers? will they actually look at the guidance in implication base and apply it to the implications? ('course they should, but....)

Kevin Power (Dec 15 2021 at 19:27):

Sounds like there is someone really pushy around here
:)

Kevin Power (Dec 15 2021 at 19:29):

I only did a quick review, and no, I didn't review general yet. I am not super concerned about the triplication, as I wouldn't anticipate it changes much - but if it does we run a risk. Perhaps we at least drop a comment in the XMLs to remind future authors where else they need to look?

Bret H (Dec 15 2021 at 19:30):

Good idea to put a comment in the XML - will do!

Bret H (Dec 15 2021 at 19:31):

So, that's quadruplication. I tried to keep it brief on General page. But could try briefer and point to the abstract Genomic Implication profile?

Kevin Power (Dec 15 2021 at 19:56):

If you are OK putting the text in genomic implication, then the text we leave in dx and tx implication XML is just a "Please see (this link) for details for AND/OR stuff..." and that is about it, that plus super simple statement in general about it along with the link to genomic implication "... for more details" should be sufficient.

Bret H (Dec 15 2021 at 19:58):

ok. I'll ping when when it's committed.

Bret H (Dec 16 2021 at 10:53):

noticed examples where evidence is conflated with clinical significance. working to address.

Bret H (Dec 16 2021 at 12:14):

sending private message

Bret H (Dec 20 2021 at 07:13):

I pushed my last changes. Anything more you would like to add @Jamie Jones ?

I suggest we fetch into the branch from master soon, as there are some changes that need to be applied to the branch which cleaned up some warnings.

Kevin Power (Dec 20 2021 at 16:52):

Do you mind bringing in the changes from master and pushing @Bret H ?

Bret H (Dec 20 2021 at 17:05):

will do

Bret H (Dec 20 2021 at 20:10):

I merged from Master to Implications branch, cured two conflicts. @Jamie Jones Do you have any figure changes to make? or other changes. After that, I think the branch is ready for Bob and Kevin to double check it against the JIRAs for intention and completion.

Bret H (Dec 20 2021 at 20:12):

(BTW locally, I get 1 error that does not block building but NO error when building through publisher to the greater world)

Bret H (Dec 20 2021 at 20:12):

https://chat.fhir.org/#narrow/stream/179297-committers.2Fnotification/topic/ig-build/near/265609934

Bret H (Dec 29 2021 at 16:27):

Any other additions? @Jamie Jones @Kevin Power shall it be pushed to master and Bob D be a reviewer? I believe he is available this week.

Kevin Power (Dec 29 2021 at 16:30):

@Bret H - There is still one JIRA that is unassigned from the list: https://jira.hl7.org/issues/?filter=16204

https://jira.hl7.org/browse/FHIR-19844?filter=16204

Bret H (Dec 30 2021 at 17:28):

@Kevin Power I will double check, but I am pretty certain we now have the values incorporated into the value set and now have examples to satisfy it. I'll assign it to myself, but I think Jaime did the value set work already, and the examples have been there for a bit. The JIRA has been around a while.

Kevin Power (Dec 30 2021 at 18:28):

OK, drop a note here once you determine it. Thanks @Bret H

Kevin Power (Dec 30 2021 at 21:28):

I merged latest QA changes from master into this branch, so if there is more to do, you should pull down changes first before making additional change.

Bret H (Dec 31 2021 at 16:24):

@Kevin Power FHIR-19844 was satisfied by Jamie's work on the value set and the examples I made for other JIRA. No JIRA remain :tada:
@Bob Dolin

Bret H (Dec 31 2021 at 16:25):

Would it make sense to add Bob D as a reviewer of the merge into master? I think it is ready to go go go : ^ )

Kevin Power (Dec 31 2021 at 20:55):

Sounds good thanks @Bret H — feel free to create the pull request and post the PR here

Bret H (Jan 03 2022 at 01:31):

Merge request: implications-updates https://github.com/HL7/genomics-reporting/pull/50

Kevin Power (Jan 03 2022 at 14:17):

Looks like our build is broken:
https://chat.fhir.org/#narrow/stream/179252-IG-creation/topic/build.2Efhir.2Eorg.20broken

Kevin Power (Jan 03 2022 at 14:18):

Not our fault, but something we will have to watch.

Kevin Power (Jan 03 2022 at 14:46):

@Jamie Jones @Bret H - I marked all the JIRAs as applied.

Bob Dolin (Jan 03 2022 at 15:27):

Is there something you want me to review?

Kevin Power (Jan 03 2022 at 15:47):

You can review the specific changes here:
https://github.com/HL7/genomics-reporting/commit/9b53a819f0f31bf226efb745ef510efebf613851

You can also review the branch here and see if the changes look good to you:
http://build.fhir.org/ig/HL7/genomics-reporting/branches/implications-updates/index.html

Bret H (Jan 03 2022 at 15:52):

for the record, on http://build.fhir.org/ig/HL7/genomics-reporting/branches/implications-updates/ValueSet-genetic-therapeutic-implications-vs.html the SCT and LNC are visible as part of the code. This is because the code system http://build.fhir.org/ig/HL7/genomics-reporting/branches/implications-updates/CodeSystem-genetic-therapeutic-implications-cs.html was built with the SCT and LNC as part of the codes.

Is there a more elegant way to represent the value set as an amalgamation of multiple code systems? LOINC and SNOMEDCT? There are other value sets which are similar in the IG. I do not see this as a problem for the merge.

Bret H (Jan 03 2022 at 15:54):

should I open a JIRA about the code system representations? To me it's really just an editing thing. A code system can represent codes however it is done, and our code system is an exemplar one, but....I can see an argument to make the codes not have the SCT and LNC.

Bret H (Jan 03 2022 at 15:56):

similarly could do some alterations on http://build.fhir.org/ig/HL7/genomics-reporting/branches/implications-updates/CodeSystem-evidence-level-example-cs.html but maybe it's more trouble than it would be worth? Given that both are code systems for Example value sets.

Kevin Power (Jan 03 2022 at 16:00):

@Bret H - Actually, can't believe I missed this, but do we even need a CodeSystem for these? Or, since the codes are all from other existing code systems, do we only need the ValueSet?

Bret H (Jan 03 2022 at 20:38):

For SCT and LNC - should not need code systems. For ClinVar stars and PGKB evidence, two code systems will be made. May believes we had a syntax problem on the therapeutic implications. She and Aly are making some changes as part of QA on our IG. Jaime was there and has more specifics, I think.

Bret H (Jan 03 2022 at 20:38):

when May and Aly are done it will be corrected : ^ )

Kevin Power (Jan 03 2022 at 20:49):

@May Terry @Aly Khalifa @Jamie Jones - If I can help with anything, let me know.

May Terry (Jan 03 2022 at 20:53):

@Kevin Power - We went over the types of changes that are needed to fix the ClinVar and PGKB code discrepancies - 2 custom code systems plus a value set which takes terms from each. Aly wanted make the change and create a PullRequest for us to review, or to reach out to me for guidance if needed. He gave the ETA of today/tomorrow early so waiting on that. @Aly Khalifa - lmk if you can't get to it by that time and I'll be happy to make that change this evening.

Aly Khalifa (Jan 03 2022 at 22:56):

@May Terry @Kevin Power @Bret H @Jamie Jones
I made the pull request, I commented out the ConditionInheritanceModeCS and GeneticTherapeuticImplicationsCS codesystems. Because they are using LOINC, SNOMED-CT and other systems codes. The codes were added directly to the corresponding valuesets. I made the two custom codesystems for ClinVar and PharmGKB containing relevant codes. The EvidenceLevelExampleVS valueset now includes codes from both systems. Alias file was updated to include HPO, GENO, and SEPIO-CG.
Please let me know if you have any comments.

May Terry (Jan 03 2022 at 23:03):

interesting. It's a PR from a fork rather than a branch from master. You okay with that, @Kevin Power ?

Kevin Power (Jan 03 2022 at 23:04):

We don't do that very often, but @Aly Khalifa may not have privs to do a branch. Its probably OK for a change like this (QA related)

May Terry (Jan 03 2022 at 23:05):

Can we give Aly permissions to branch from now on? He is officially a code contributor. :-)

Aly Khalifa (Jan 03 2022 at 23:06):

:)

Kevin Power (Jan 03 2022 at 23:06):

You have to ask Lloyd for that :smile:

Kevin Power (Jan 03 2022 at 23:07):

One preference - do you mind just deleting lines instead of commenting out. The GitHub diff makes is clearer when something is deleted (at least in my experience)

Aly Khalifa (Jan 03 2022 at 23:08):

sure, sorry for that

May Terry (Jan 03 2022 at 23:08):

k. we'll sort that out next week. In the meantime, I cloned Aly's forked code so I can run it locally.

Kevin Power (Jan 03 2022 at 23:09):

No worries @Aly Khalifa - many thanks for doing this.

I too have cloned locally and am taking a look.

Kevin Power (Jan 03 2022 at 23:10):

@Aly Khalifa - Did you do a local build?

Aly Khalifa (Jan 03 2022 at 23:10):

I removed them

May Terry (Jan 03 2022 at 23:10):

yeh...IG Publisher is failing on that codebase

May Terry (Jan 03 2022 at 23:10):

Error: java.lang.Exception: Unable to find the source file for CodeSystem/condition-inheritance-mode-cs: not specified

Aly Khalifa (Jan 03 2022 at 23:11):

@Kevin Power - no I did not

Kevin Power (Jan 03 2022 at 23:11):

yea, these have to be removed from the IG file as well. (found in input/genomics-reporting.xml)

Aly Khalifa (Jan 03 2022 at 23:12):

@May Terry - which file is producing this error.

Kevin Power (Jan 03 2022 at 23:12):

(which, btw, not every artifact is in that file)

Kevin Power (Jan 03 2022 at 23:13):

Aly Khalifa said:

May Terry - which file is producing this error.

input/genomics-reporting.xml)

Kevin Power (Jan 03 2022 at 23:16):

@Aly Khalifa - If you get me write access to your Fork, I can fix this problem.

Kevin Power (Jan 03 2022 at 23:17):

https://github.com/sirkyp

Aly Khalifa (Jan 03 2022 at 23:21):

@Kevin Power - I've just did

Aly Khalifa (Jan 03 2022 at 23:21):

what changes need to be done to this file?

Kevin Power (Jan 03 2022 at 23:23):

Need to remove these lines:

    <resource>
      <reference>
        <reference value="CodeSystem/condition-inheritance-mode-cs"/>
      </reference>
      <name value="Condition Inheritance Pattern Codes"/>
      <description value="A set of terms that describe the transmission pattern of a condition in a pedigree"/>
      <groupingId value="codesystems"/>
    </resource>
    <resource>
      <reference>
        <reference value="CodeSystem/genetic-therapeutic-implications-cs"/>
      </reference>
      <name value="Genetic Therapeutic Implication Codes"/>
      <description value="A set of terms that describe a predicted ramification based on the presence of associated molecular finding(s)."/>
      <groupingId value="codesystems"/>
    </resource>

Aly Khalifa (Jan 03 2022 at 23:24):

ok, I'll remove them now

Kevin Power (Jan 03 2022 at 23:24):

But, you need to run a build afterwards to make sure no other problems are occuring.

Kevin Power (Jan 03 2022 at 23:25):

https://confluence.hl7.org/display/FHIR/IG+Publisher+Documentation

Kevin Power (Jan 03 2022 at 23:25):

If you have never setup the build process, it can take a bit the first time.

Aly Khalifa (Jan 03 2022 at 23:27):

Exactly, I do not have the setup and this would be my first time

Aly Khalifa (Jan 03 2022 at 23:28):

@Kevin Power - However, I removed the lines

Kevin Power (Jan 03 2022 at 23:28):

Looks like there is an example that uses the condition-inheritance-mode-cs

Kevin Power (Jan 03 2022 at 23:29):

I have to drop off for a while, but I can pick this up tomorrow. @Aly Khalifa - You can either give me write privs to your fork, or I can just take your work and create a branch. But I will have to do that tomorrow.

Aly Khalifa (Jan 03 2022 at 23:30):

I did send you the invitation through GitHub

Aly Khalifa (Jan 03 2022 at 23:31):

few minutes ago, did you get it?

Aly Khalifa (Jan 03 2022 at 23:34):

I sent it again through Zulip. bye

May Terry (Jan 04 2022 at 00:19):

@Aly Khalifa - looks good on my end. Your latest commit fixed both the IG Publisher error and also displays the code systems and value sets correctly as we discussed earlier. Good job! :smile:

Kevin Power (Jan 04 2022 at 04:28):

@Aly Khalifa - I got it and will look tomorrow

Aly Khalifa (Jan 04 2022 at 07:00):

@May Terry - Thanks, this is great :smile:

Aly Khalifa (Jan 04 2022 at 07:56):

file: input\examples\obs-idh-ex.xml contains a coding that uses condition-inheritance-mode-cs as part of valueCodeableConcept element

Aly Khalifa (Jan 04 2022 at 07:58):

<valueCodeableConcept>
  <coding>
    <system value="http://hl7.org/fhir/uv/genomics-reporting/CodeSystem/condition-inheritance-mode-cs"/>
    <code value="HP:0000007"/>
    <display value="Autosomal recessive inheritance"/>
  </coding>
</valueCodeableConcept>

Aly Khalifa (Jan 04 2022 at 08:13):

I'll change that to reflect the updated system and code, i.e., HPO and code "0000007".
For consistence, I'll also change the used URL and abbreviation/alias to reflect those mentioned on confluence pages:

• Code System Metadata Records Index: https://confluence.hl7.org/pages/viewpage.action?pageId=79497166
• Human Phenotype Ontology page: https://confluence.hl7.org/display/TA/HPO+The+Human+Phenotype+Ontology
  I used the HPO URL of EMBL-EBI while Confluence lists the Jackson lab URL. Also, HP was used for abbreviation while Confluence lists HPO as the one. We could also use the OID.

Aly Khalifa (Jan 04 2022 at 08:23):

I pushed these changes to GitHub.

Kevin Power (Jan 04 2022 at 14:55):

I did some builds and testing, and pushed up a few other changes to Aly's fork. @May Terry - Do you have a bit to review?

May Terry (Jan 04 2022 at 14:56):

Sure. I can review.

Bret H (Jan 04 2022 at 14:59):

@Aly Khalifa what URI or URL did you use for GENO, HPO, SEPIO-CG? Saw the comment earlier. Are all the URLs from the confluence page?

Aly Khalifa (Jan 04 2022 at 16:05):

@Bret H -
GENO = https://www.ebi.ac.uk/ols/ontologies/geno
HPO = https://hpo.jax.org/app/
// The SEPIO-CG link is according to https://dataexchange.clinicalgenome.org/interpretation/terminologies/ but not working
SEPIO-CG = http://purl.obolibrary.org/obo/sepio-clingen

Bret H (Jan 04 2022 at 16:18):

Much appreciated! Those look good to me. We should connect with @Joel Schneider if we would like them to be official with the Terminology WG. Joel's our vocabulary officer.

Bret H (Jan 04 2022 at 16:19):

I'm not sure myself how important it is to have an HL7 recognized URL/URI. But it would mean that developers could see them on the HL7 vocabulary pages.

Kevin Power (Jan 04 2022 at 16:25):

I will quickly add that I edited them: HTTPS (I deleted the S as that is common when defining these)

Aly Khalifa (Jan 04 2022 at 16:59):

@May Terry - I sent you a GitHub invitation to our fork

Kevin Power (Jan 04 2022 at 23:14):

@May Terry - If you approve it, I will go ahead and merge it.

Joel Schneider (Jan 05 2022 at 00:50):

For cases where we need an official code system identifier for an external code system, it would be best to follow the HTA process described here:
https://confluence.hl7.org/display/TA/Validating+and+Requesting+Code+System+Identifiers+for+an+External+Code+System

(The HTA is not the same as the Vocabulary WG, although there may be some overlap in personnel.)

May Terry (Jan 05 2022 at 15:42):

sounds good @Kevin Power . just cloned and reviewing now.

May Terry (Jan 05 2022 at 15:54):

Makes sense @Joel Schneider - Can you submit the HTA request for any new code systems in parallel? That way we're not blocked by a system that, while necessary, can be a delayed because of their backlog.

Joel Schneider said:

For cases where we need an official code system identifier for an external code system, it would be best to follow the HTA process described here:
https://confluence.hl7.org/display/TA/Validating+and+Requesting+Code+System+Identifiers+for+an+External+Code+System

(The HTA is not the same as the Vocabulary WG, although there may be some overlap in personnel.)

Joel Schneider (Jan 05 2022 at 15:57):

@May Terry My understanding is that a separate HTA request should be created for each new (external) code system identifier.

May Terry (Jan 05 2022 at 16:02):

Right - I was just hoping that you can do this for us from a tasking perspective, even if the codes themselves are not officially in there. If not then @Kevin Power or @Bret H can handle it as well - looks like we need to submit the request for PharmGKB.
This usage follows our pattern of TBD LOINC codes because they need to go through a process of addition but we don't want to be blocked by the stewards who own those knowledgebases.

Kevin Power (Jan 05 2022 at 16:36):

Perhaps we should start a CG JIRA first, and start out with our group listing out and proposing a list of code systems that we should request?

Kevin Power (Jan 05 2022 at 18:33):

@Aly Khalifa - Would you be willing to create a JIRA for CG and gather feedback from the community on code systems we need to add?

May Terry (Jan 05 2022 at 18:36):

@Kevin Power - do you still want me to approve the last commit for this PR even if the other code systems are not displaying in the artifacts list? or do you want to us to troubleshoot/fix it before we merge?

Kevin Power (Jan 05 2022 at 18:36):

@May Terry - Dang, I had forgot about that one. :frown: I have a bit, I will take a quick look.

May Terry (Jan 05 2022 at 18:38):

and @Aly Khalifa - per the latest PullRequest changes, the code systems we have that might need to submit (one per code system) in JIRA for HTA have the following canonical URLs...

• http://www.ebi.ac.uk/ols/ontologies/geno
• http://hpo.jax.org/app/
• http://purl.obolibrary.org/obo/sepio-clingen
• http://www.pharmgkb.org

Aly Khalifa (Jan 05 2022 at 18:48):

@Kevin Power @May Terry - I have not done this before. What is the process for creating a JIRA for CG? I may need time to know how to do that.

Kevin Power (Jan 05 2022 at 18:49):

Request an account on jira.hl7.org @Aly Khalifa

Kevin Power (Jan 05 2022 at 18:49):

@May Terry - I see those code systems on the artifacts page (just in their own section, given what I talked about yesterday on the call)

image.png

May Terry (Jan 05 2022 at 18:50):

ah...k. I thought we were going to see about combining them all in one section. but if we're okay with this then I don't have any other glaring issues noticed. I'll approve. Thanks! :-)

Kevin Power (Jan 05 2022 at 18:52):

We have several other artifacts that are already in this boat :smile:

May Terry (Jan 05 2022 at 18:53):

true. anyway, I just approved.

Kevin Power (Jan 05 2022 at 19:01):

Merged! Many thanks @Aly Khalifa and @May Terry

Aly Khalifa (Jan 05 2022 at 19:12):

I've just made the account, and I'll check the documentation for how to submit these requests.

Kevin Power (Jan 05 2022 at 19:16):

go here:
https://jira.hl7.org/projects/FHIR/issues/FHIR-32581?filter=allopenissues

Click the Create button at the top of the screen.

Aly Khalifa (Jan 05 2022 at 19:48):

@Kevin Power @May Terry - please check this one and if it looks good I'll do the same for each of the three remaining codying systems: https://jira.hl7.org/projects/FHIR/issues/FHIR-34917

Kevin Power (Jan 05 2022 at 22:06):

Looks like a good start to me. You might take a look at the link from Joel and see if there is anything else we would need to provide in our request to get it created, and see if we can start documenting those things on our JIRA?

Bret H (Jan 06 2022 at 14:17):

Other than the IG, I thought we had a confluence page with relevant vocab systems?

Kevin Power (Jan 06 2022 at 14:21):

CG specific ones?

Bret H (Jan 06 2022 at 14:23):

did a quick look. did not see the page I recall, must just be me. No worries.

Aly Khalifa (Jan 06 2022 at 15:09):

@Bret H - do you mean this one https://confluence.hl7.org/pages/viewpage.action?pageId=79497166 ?

Aly Khalifa (Jan 06 2022 at 15:22):

Kevin Power said:

Looks like a good start to me. You might take a look at the link from Joel and see if there is anything else we would need to provide in our request to get it created, and see if we can start documenting those things on our JIRA?

Ok, I'll do that

Bret H (Jan 06 2022 at 15:23):

nope. it was specific to the wg. no worries, its not important.

Aly Khalifa (Jan 07 2022 at 07:26):

May Terry said:

and Aly Khalifa - per the latest PullRequest changes, the code systems we have that might need to submit (one per code system) in JIRA for HTA have the following canonical URLs...

• http://www.ebi.ac.uk/ols/ontologies/geno
• http://hpo.jax.org/app/
• http://purl.obolibrary.org/obo/sepio-clingen
• http://www.pharmgkb.org

I would add the clinvar Review Status: http://www.ncbi.nlm.nih.gov/clinvar/docs/review_status/ which us currently used as custom code system similar to PharmGKB clinical annotation levels of evidence which would have the following URL of pharmgkb including the codes: http://www.pharmgkb.org/page/clinAnnLevels

May Terry (Jan 07 2022 at 15:35):

Good catch.
I thought ClinVar was already recognized by HTA since it's been listed here (http://hl7.org/fhir/terminologies-systems.html) for a while, but it appears that there is again a misalignment between what's listed in HL7 FHIR external terminologies and the HTA list here (https://confluence.hl7.org/pages/viewpage.action?pageId=79497166).

Aly Khalifa said:
I would add the clinvar Review Status: http://www.ncbi.nlm.nih.gov/clinvar/docs/review_status/ which us currently used as custom code system similar to PharmGKB clinical annotation levels of evidence which would have the following URL of pharmgkb including the codes: http://www.pharmgkb.org/page/clinAnnLevels

Bret H (Jan 07 2022 at 16:43):

Yep. That's one reason I created a CS that had the info from CLinVar and PharmGKB as a toy-example. But pursing the HTA listing and building it like it is now will reduce barriers for real-world use of the codes. Great work team!

Bret H (Jan 07 2022 at 16:49):

@Aly Khalifa @May Terry On http://build.fhir.org/ig/HL7/genomics-reporting/CodeSystem-pharmgkb-evidence-level-custom-cs.html#root is this correct: image.png

Should the title be: CodeSystem: PharmGKB Evidence Level Example codes

Change here too? Title: Evidence Level Example codes

Aly Khalifa (Jan 07 2022 at 17:50):

@Bret H I've just changed it in the fork.

Aly Khalifa (Jan 07 2022 at 19:38):

@May Terry @Kevin Power @Bret H - As per JIRA HTA guidance I checked various sources for the code systems we have. Except for HPO, they are not there. More details are in the attached file.
We already used the listed URL of HPO at Confluence HTA page. There is a question about pharmGKB and Clinvar, they are listed in on FHIR Terminology page and HL7 OID registry, however, I suppose these listing are for variations, allies, and drug labels, than Review Status and Level of Evidence (LOE) coding schemas which is our use case. Do we add ClinVar-Review Status and PharmGKB-LOE as seperate codying systems besides ClinVar variants/allies and PharmGKB drug labels and annotations? New-code-systems-for-HTA_20220107_AK.xlsx

Bret H (Jan 07 2022 at 23:14):

correct. the listing is more at the source system level (and geared towards variant or knowledge base info) rather than levels of evidence. Luckily, there are specific urls to each representation of the 'code systems' for the evidence values from the two resources (PharmGKB and ClinVar). I would be supportive of adding new code systems for the evidence levels, as the current definitions are not focused on the resource as a whole. It is different than referring to SNOMED CT , which is expressly an ontology, or Sequence Ontology.

Aly Khalifa (Jan 10 2022 at 15:35):

JIRA HTA issues
-GENO https://jira.hl7.org/browse/HTA-67
-SEPIO-CG https://jira.hl7.org/browse/HTA-68
-CLINVAR-RS https://jira.hl7.org/browse/HTA-69
-PGKB-ClinAnnLevels https://jira.hl7.org/browse/HTA-70

Aly Khalifa (Jan 10 2022 at 16:10):

JIRA-Genomics Reporting (FHIR) issues:

• GNEO https://jira.hl7.org/browse/FHIR-34917
• SEPIO-CG https://jira.hl7.org/browse/FHIR-35298
• CLINVAR-RS https://jira.hl7.org/browse/FHIR-35303
• PGKB-ClinAnnLevels https://jira.hl7.org/browse/FHIR-35306

Aly Khalifa (Jan 10 2022 at 16:43):

@Joel Schneider @Jamie Jones - I added above the JIRA issues for HTA

Jamie Jones (Jan 10 2022 at 17:04):

@Aly Khalifa here is the previous code system worksheet I found: https://docs.google.com/spreadsheets/d/14sIVVED2rI0Ab2ZQ3oV6v8njzX-sRezkIWcgkkepUUg/edit#gid=1230040956.
If you get a chance, it would be great if you could review this for other issues we may need to resurface. Thanks!

Kevin Power (Jan 10 2022 at 21:26):

Not sure what others think here, but I was suggesting that we have @Aly Khalifa log the CG JIRA(s) to indicate which Code System(s) we would log to log to HTA, and have the CG community review and approved before logging the HTA JIRAs. However, since Aly already logged them to HTA, and honestly it seems like a reasonable list to me, I would propose we just cancel the CG FHIR issues. Anyone disagree?

Aly Khalifa (Jan 11 2022 at 16:37):

Jamie Jones said:

Aly Khalifa here is the previous code system worksheet I found: https://docs.google.com/spreadsheets/d/14sIVVED2rI0Ab2ZQ3oV6v8njzX-sRezkIWcgkkepUUg/edit#gid=1230040956.
If you get a chance, it would be great if you could review this for other issues we may need to resurface. Thanks!

Sure, I'll check the included URLs same as I did for aforementioned ones.

Aly Khalifa (Jan 13 2022 at 14:42):

@Jamie Jones - I checked the first 13 entries shaded in green. All of them are on https://www.hl7.org/fhir/terminologies-systems.html
As per HTA policy "Validating and Requesting Code System Identifiers for an External Code System" we do not need to make a request to add them. We may only use the provided URLs at https://www.hl7.org/fhir/terminologies-systems.html. However, I noticed that there are more than on OID for some systems. We may need to double check this. More details are in the attached Excel file. New-code-systems-for-HTA_20220111_AK.xlsx

Joel Schneider (Jan 13 2022 at 20:27):

HL7's goal is to have terminology.hl7.org (THO) become the authoritative source of information about code systems cited in HL7 published artifacts. E.g. https://terminology.hl7.org/external_terminologies.html

For some code systems, we (CG) might also want to collaborate with the Vocabulary WG to develop "Using X with FHIR" guidance as part of the FHIR spec.

Incidentally, I was able to attend the HTA call this morning, and we made some headway on the older JIRA tickets.

Patrick Werner (Feb 01 2022 at 19:03):

Kevin Power said:

Not sure what others think here, but I was suggesting that we have Aly Khalifa log the CG JIRA(s) to indicate which Code System(s) we would log to log to HTA, and have the CG community review and approved before logging the HTA JIRAs. However, since Aly already logged them to HTA, and honestly it seems like a reasonable list to me, I would propose we just cancel the CG FHIR issues. Anyone disagree?

i think we should keep our issues as reminders to apply the new/approved urls after HTAs decision.

Kevin Power (Feb 01 2022 at 20:21):

Good call @Patrick Werner

Last updated: Apr 12 2022 at 19:14 UTC