This wiki has undergone a migration to Confluence found Here
<meta name="googlebot" content="noindex">

Difference between revisions of "201901 Clinical Genomics"

From HL7Wiki
Jump to navigation Jump to search
 
(50 intermediate revisions by 6 users not shown)
Line 8: Line 8:
 
The Zulip Chat will be used to coordinate among participants during the connectathon:
 
The Zulip Chat will be used to coordinate among participants during the connectathon:
  
[https://chat.fhir.org/#narrow/stream/58-connectathon-mgmt/subject/Clinical.20Genomics.20Track Zulip Chat]
+
[https://chat.fhir.org/#narrow/stream/179197-genomics/topic/Jan.202019.20Connectathon]
 
 
  
 
==Submitting WG/Project/Implementer Group==
 
==Submitting WG/Project/Implementer Group==
Line 19: Line 18:
 
Genomic data are of increasing importance to clinical care and secondary analysis. After initial Feedback from the first ballot of the Clinical Genomics Implementation Guide, considerable updates have been made, and the recently balloted IG and profiles should be tested in the Connectathon for more feedback and trial use.
 
Genomic data are of increasing importance to clinical care and secondary analysis. After initial Feedback from the first ballot of the Clinical Genomics Implementation Guide, considerable updates have been made, and the recently balloted IG and profiles should be tested in the Connectathon for more feedback and trial use.
  
==Proposed Track Lead==
+
==Track Lead==
 
<!-- Name, email and Skype id of individual who will coordinate the track at the connectathon -->
 
<!-- Name, email and Skype id of individual who will coordinate the track at the connectathon -->
  [mailto:gilusa@gmail.com Gil Alterovitz], [mailto:patrick.werner@molit.eu Patrick Werner]
+
  [mailto:gilusa@gmail.com Gil Alterovitz]
  
 
==Expected participants==
 
==Expected participants==
 
<!-- List of the individuals and/or organizations that have indicated a desire to attend the connectathon and implement this track -->
 
<!-- List of the individuals and/or organizations that have indicated a desire to attend the connectathon and implement this track -->
Gil Alterovitz, James Jones, Kevin Power ...
+
Gil Alterovitz, Patrick Werner, James Jones, Kevin Power, Lloyd Mckenzie, Bret Heale, Bob Millius, Joel Schneider, Bob Dolin, Bob Freimuth, Stephanie Garcia, Apurva Dharia, Matt Tiller, Weill Cornell Medicine, Utah Newborn Screening Program, NMDP.
 +
 
  
 
==Roles==
 
==Roles==
Line 31: Line 31:
 
===FHIR Client===
 
===FHIR Client===
 
<!-- Provide a description of the capabilities this role will have within the connectathon -->
 
<!-- Provide a description of the capabilities this role will have within the connectathon -->
Support the sending of the Sequence resource/genetics profiles operations: create, history, read, search and update.
+
Support the sending of the MolecularSequence resource and genetics profiles operations: create, history, read, search and update.
  
 
===FHIR Server===
 
===FHIR Server===
 
<!-- Provide a description of the capabilities this role will have within the connectathon -->
 
<!-- Provide a description of the capabilities this role will have within the connectathon -->
Support the receiving and processing of the Sequence resource/genetics profiles operations: create, history, read, search and update.
+
Support the receiving and processing of the MolecularSequence resource/genetics profiles operations: create, history, read, search and update.
  
 
==Scenarios==
 
==Scenarios==
 
<!-- What will be the actions performed by participants? -->
 
<!-- What will be the actions performed by participants? -->
 +
Scenarios 2-6 are taken from use cases in the [http://www.hl7.org/implement/standards/product_brief.cfm?product_id=479 HL7 Domain Analysis Model (DAM): Clinical Genomics, Release 1, July 2018 Informative Publication].  See links below each scenario for relevant DAM excerpts.
 +
A couple new use cases may be considered as approved by the CG workgroup and time permits.
 +
 +
===Scenario 1 Register a New MolecularSequence, DescribedVariant, and GenomicsReport ===
 +
:'''Action''': (FHIR Client) (FHIR Client) Create 1 or all of: a MolecularSequence instance, a (DescribedVariant-profiled) Observation instance, and/or a (GenomicsReport-profiled) DiagnosticReport instance to represent genetics data and interpretations (DNA variant, RNA sequence, structural variant, etc). <!--Who does what?  (Use the role names listed above when referring to the participants -->
 +
:'''Precondition''': The instances do not exist in server prior to action. <!-- What setup is required prior to executing this step? -->
 +
:'''Success Criteria''': Instances created correctly on server and in the desired format. <!-- How will the participants know if the test was successful? -->
 +
:'''Bonus point''': GenomicsReport resources tailored to individual use cases can be sent and received.<!-- Any additional complexity to make the scenario more challenging -->
 +
<!-- Provide a description of each task -->
  
===Register a New Sequence and Observation===
 
===Register a New Genomics Report===
 
===Scenario 1 PGx (placeholder)===
 
 
===Scenario 2 Clinical Sequencing - Germline Testing===
 
===Scenario 2 Clinical Sequencing - Germline Testing===
 +
[https://docs.google.com/document/d/1-PULyIwtpCZmsANpafzYA2CKJ1ZMiIysdNnFoGf2-KY/edit#heading=h.gjdgxs?usp=sharing/ More on Scenario 2]
 +
:'''Action''': (FHIR Client) Search target genetic-profile-based observation with given patient ID. This target observation has a component - Genomic Source Class (code={"coding":[{"system":"http://loinc.org","code":"48002-0"}]}, value="LA6683-2") corresponding to Germline.
 +
:'''Precondition''': Relevant patient and observations have been created
 +
:'''Success Criteria''': A bundle of genetics observations from germline analysis of that patient are returned.<!-- How will the participants know if the test was successful? -->
 +
:'''Bonus point''': More parameters can be added for searching<!-- Any additional complexity to make the scenario more challenging -->
 +
 
===Scenario 3 Family Member History===
 
===Scenario 3 Family Member History===
 +
[https://docs.google.com/document/d/1-PULyIwtpCZmsANpafzYA2CKJ1ZMiIysdNnFoGf2-KY/edit#heading=h.3dy6vkm?usp=sharing/ More on Scenario 3]
 +
:'''Action''': (FHIR Client) 1. Search the genomics reports of a given patient.  Get the reference id of this patient’s familymemberhistory from the supportingInfo extension - Family History (http://hl7.org/fhir/uv/genomics-reporting/dr-supportingInfo.html). 2. Retrieve this FamilyMemberHistory instance and get the reference id of Observation instances from extension - Genetic markers, ethnicity, etc. (url: http://hl7.org/fhir/StructureDefinition/family-member-history-genetics-observation). 3. Retrieve the target Observation instance.
 +
:'''Precondition''': Relevant diagnostic report, observation, family member history have been created.
 +
:'''Success Criteria''': The target Observation instances of the given patient's family member are returned.<!-- How will the participants know if the test was successful? -->
 +
:'''Bonus point''': More parameters can be added for searching<!-- Any additional complexity to make the scenario more challenging -->
 +
 
===Scenario 4 Clinical and Research Data Warehouses===
 
===Scenario 4 Clinical and Research Data Warehouses===
 +
[https://docs.google.com/document/d/1-PULyIwtpCZmsANpafzYA2CKJ1ZMiIysdNnFoGf2-KY/edit#heading=h.4d34og8?usp=sharing/ More on Scenario 4]
 +
:'''Action''': (FHIR Client) Search for all observations with a given described variant (e.g. rs587778247)
 +
:'''Precondition''': Observation instances with this variant have been created
 +
:'''Success Criteria''': A bundle of describedVariant observations, where the code value of component - dbSNP Id ({"coding":[{"system":"http://loinc.org","code":"81255-2"}]} is rs587778247. <!-- How will the participants know if the test was successful? -->
 +
:'''Bonus point''': More parameters can be added for searching<!-- Any additional complexity to make the scenario more challenging -->
 +
 
===Scenario 5 HLA Typing===
 
===Scenario 5 HLA Typing===
===Scenario 6 Comprehensive Pathology Report===
+
[https://docs.google.com/document/d/1-PULyIwtpCZmsANpafzYA2CKJ1ZMiIysdNnFoGf2-KY/edit#heading=h.17dp8vu?usp=sharing/ More on Scenario 5]
 +
:'''Action''': (FHIR Client) Create an HLA genotyping genomics report.
 +
:'''Precondition''': This genomicsReport instance does not exist in service prior to action.
 +
:'''Success Criteria''': The genomics report carrying HLA information is created correctly on server and in the desired format.
 +
:'''Bonus point''': Extensions can be added<!-- Any additional complexity to make the scenario more challenging -->
  
===PGx CDS Scenario===
+
===Scenario 6 Species Identification===
Please see detailed storyboard here: https://docs.google.com/document/d/1VqPrf6aaOB8RhSbAq2JYDxZYbDm-RT9DxA0u6-RxYo0/edit  
+
[https://docs.google.com/document/d/1-PULyIwtpCZmsANpafzYA2CKJ1ZMiIysdNnFoGf2-KY/edit#heading=h.tyjcwt?usp=sharing/ More on Scenario 6]
 +
:'''Action''': (FHIR Client) Search for sequences from human (Homo sapiens).
 +
:'''Precondition''': Sequence(s) have been created where the species is defined as human (Homo sapiens).
 +
:'''Success Criteria''': A bundle of Sequence instances where the value of the species element is Homo sapiens.
  
A swim-lane overview of the scenario is here:
+
===Scenario 7 Comprehensive Pathology Report===
 +
:'''Action''': (FHIR Client) Create a comprehensive pathology report which includes genetic information for a patient.  A comprehensive pathology report integrates pertinent information gathered from various methods (e.g. morphology, immunohistochemistry, flow cytometry, cytogenetics, fluorescence ''in situ'' hybridization [FISH], and molecular testing [e.g., NGS]). The genomicsReport profile has the capability to support results with simple or complex genetics observations.
 +
:'''Precondition''': This genomics report has never been created
 +
:'''Success Criteria''': Report created correctly on server and in the desired format.
 +
:'''Bonus point''': Extensions can be added.
  
[[File:PharmacogenomicsCDS.png|600px]]
+
===Scenario 8 Sequence quality===
 +
:'''Action''': (FHIR Client) Get the quality of the sequence under consideration.  In the sequencing reads, each base is assigned with a quality score generated by the sequencer, which represents the confidence of a base call. Base quality is a critical factor for accurate variant detection in the downstream analysis.
 +
:'''Precondition''': This sequence instance has been created
 +
:'''Success Criteria''': Target sequence instance is returned.
 +
:'''Bonus point''': More parameters can be added for searching
  
Additional Roles
+
===Scenario 9 PGx CDS===
1. PGx CDS Service: Interact with EHR (via CDS Hooks), OMIC Data Store (for genomoic observations), and Terminology Server (for medication value sets).
+
This scenario demonstrates an implementation of CPIC Pharmacogenomic Guidelines [https://cpicpgx.org/] based on FHIR, CDS Hooks [https://cds-hooks.org/], and genomics results formatted per our latest HL7 FHIR Clinical Genomics Reporting guide [http://hl7.org/fhir/uv/genomics-reporting/index.html]
  
2. OMIC Data Store: Store genomic observations
+
:'''Action''': In this scenario, a medication order arises in the EHR (which is also a CDS Hooks client), triggering a notification to a PGx CDS Service (which is also a CDS Hooks server). The "medication-prescribe" hook also provides the PGX CDS Service with a FHIR Medication resource that includes an RxNorm code for the ordered drug. Having been triggered by the “medication-prescribe” CDS hook, the PGx CDS Service now executes a decision support rule that [1] determines if the ordered drug has a known gene interaction; [2] determines, where there is a known drug-gene interaction, whether or not the patient has genetic test results on file; [3] determines, where there are genetic test results on file, if the patient has an interacting genotype; [4] determines, where there are not genetic test results on file, if the patient needs pre-testing. The PGx CDS Service returns a CDS hooks “suggestion card” back to the EHR, with appropriate recommendations.
  
3. EHR: Order entry of a medication triggers a “medication-prescribe” CDS hook which invokes the PGx CDS Service and provides it with a FHIR MedicationOrder instance.
+
:'''Precondition''': Genomic observation(s) are available for some patients.
  
4. Terminology Service: The PGx CDS Service will ping the terminology server to see if the ordered drug is in an “Actionable PGx Drug” value set.
+
:'''Success Criteria''': Retrieve relevant observations (variants, haplotype, genotype, etc) on a pharmacogene where they exist.
  
:'''Action''': [1] (EHR) Manually enter an order for "Imuran 50mg 1 tablet twice a day by mouth" into order entry system. This triggers a “medication-prescribe” CDS hook with a hook context of MedicationOrder which includes RxNorm code 197388 for azathioprine 50mg oral tablet; [2] (PGx CDS Service) Having been triggered by the “medication-prescribe” CDS hook, the PGx CDS Service now executes a decision support rule that [1] determines if the ordered drug has a known gene interaction; [2] determines, where there is a known drug-gene interaction, whether or not the patient has genetic test results on file; [3] determines, where there are genetic test results on file, if the patient has an interacting genotype; [4] determines, where there are not genetic test results on file, if the patient needs pre-testing. PGx CDS Service returns a CDS hooks “information card” back to the EHR, with appropriate recommendations.
+
:'''Bonus point''': Correctly identify that there are no observations for the gene in question.
  
:'''Precondition''': Genomic observation(s) are available for some patients.
+
A swim-lane overview of the scenario is here ('GACS' is a Genomic Archiving and Communication System, used to store genomic test results):
 +
 
 +
[[File:PharmacogenomicsCDS2.png|600px]]
 +
 
 +
===Scenario 10: Implementing Genomics Diagnostic Reports===
  
:'''Success Criteria''': Retrieve relevant observations (on TPMT gene, on allelic state) where they exist.
+
Goal: To learn what types of local modifications sites may require when adopting the Clinical Genomics FHIR resources and profiles, and to uncover gaps in the specification to meet emerging use cases.
  
:'''Bonus point''': Search OMIC data store for different potential representations.
+
:'''Action''': Each client sends a locally-customized FHIR Genomics Diagnostic Report (GDR) to a common receiving server.
 +
:'''Preconditions''': Preconfigured FHIR server without customized resources present, GDRs created by each site.
 +
:'''Success Criteria''': Receiving server accepts the GDR post and all data elements contained within it.
 +
:'''Bonus point''': Successful exchange of GDRs between servers, and actionable feedback/enhancement requests.
  
:'''Reference''': Azathioprine Pharmacogenomics - CPIC recommendations [https://cpicpgx.org/guidelines/guideline-for-thiopurines-and-tpmt/]
+
This is scenario was designed by the ONC Sync for Genes Phase 2 pilot site participants but it is widely applicable to clinical genomics use cases. For more information, contact scenario coordinator Bob Freimuth.
  
===What else?===
+
'''Final Genomics Diagnostics Reports for Scenario 10'''
 +
:'''Utah Newborn Screening Program:''' [[File:Final_S4GP2_Connectathon_Utah_NBS.zip]]
 +
:'''Weill Cornell Medical Center:''' [[File:Completed_Cornell_CaT_Files.zip]]
 +
:'''NMDP:''' [[File:NMDP_Original_and_Final_Files.zip]]
  
 
==TestScript(s)==
 
==TestScript(s)==
Line 81: Line 130:
 
These should be committed to SVN under trunk/connectathons/[connectathon]
 
These should be committed to SVN under trunk/connectathons/[connectathon]
 
-->
 
-->
 +
The supporting TestScripts and corresponding fixtures have been committed to the FHIR documents Github repository at: https://github.com/FHIR/documents/tree/master/connectathons/NewOrleansJan2018/Connectathon17/Genomics
 +
  
 
==Security and Privacy Considerations==
 
==Security and Privacy Considerations==

Latest revision as of 23:55, 20 March 2019


Clinical Genomics Track

Zulip Chat

The Zulip Chat will be used to coordinate among participants during the connectathon:

[1]

Submitting WG/Project/Implementer Group

Justification

Genomic data are of increasing importance to clinical care and secondary analysis. After initial Feedback from the first ballot of the Clinical Genomics Implementation Guide, considerable updates have been made, and the recently balloted IG and profiles should be tested in the Connectathon for more feedback and trial use.

Track Lead

Gil Alterovitz

Expected participants

Gil Alterovitz, Patrick Werner, James Jones, Kevin Power, Lloyd Mckenzie, Bret Heale, Bob Millius, Joel Schneider, Bob Dolin, Bob Freimuth, Stephanie Garcia, Apurva Dharia, Matt Tiller, Weill Cornell Medicine, Utah Newborn Screening Program, NMDP.


Roles

FHIR Client

Support the sending of the MolecularSequence resource and genetics profiles operations: create, history, read, search and update.

FHIR Server

Support the receiving and processing of the MolecularSequence resource/genetics profiles operations: create, history, read, search and update.

Scenarios

Scenarios 2-6 are taken from use cases in the HL7 Domain Analysis Model (DAM): Clinical Genomics, Release 1, July 2018 Informative Publication. See links below each scenario for relevant DAM excerpts. A couple new use cases may be considered as approved by the CG workgroup and time permits.

Scenario 1 Register a New MolecularSequence, DescribedVariant, and GenomicsReport

Action: (FHIR Client) (FHIR Client) Create 1 or all of: a MolecularSequence instance, a (DescribedVariant-profiled) Observation instance, and/or a (GenomicsReport-profiled) DiagnosticReport instance to represent genetics data and interpretations (DNA variant, RNA sequence, structural variant, etc).
Precondition: The instances do not exist in server prior to action.
Success Criteria: Instances created correctly on server and in the desired format.
Bonus point: GenomicsReport resources tailored to individual use cases can be sent and received.

Scenario 2 Clinical Sequencing - Germline Testing

More on Scenario 2

Action: (FHIR Client) Search target genetic-profile-based observation with given patient ID. This target observation has a component - Genomic Source Class (code={"coding":[{"system":"http://loinc.org","code":"48002-0"}]}, value="LA6683-2") corresponding to Germline.
Precondition: Relevant patient and observations have been created
Success Criteria: A bundle of genetics observations from germline analysis of that patient are returned.
Bonus point: More parameters can be added for searching

Scenario 3 Family Member History

More on Scenario 3

Action: (FHIR Client) 1. Search the genomics reports of a given patient. Get the reference id of this patient’s familymemberhistory from the supportingInfo extension - Family History (http://hl7.org/fhir/uv/genomics-reporting/dr-supportingInfo.html). 2. Retrieve this FamilyMemberHistory instance and get the reference id of Observation instances from extension - Genetic markers, ethnicity, etc. (url: http://hl7.org/fhir/StructureDefinition/family-member-history-genetics-observation). 3. Retrieve the target Observation instance.
Precondition: Relevant diagnostic report, observation, family member history have been created.
Success Criteria: The target Observation instances of the given patient's family member are returned.
Bonus point: More parameters can be added for searching

Scenario 4 Clinical and Research Data Warehouses

More on Scenario 4

Action: (FHIR Client) Search for all observations with a given described variant (e.g. rs587778247)
Precondition: Observation instances with this variant have been created
Success Criteria: A bundle of describedVariant observations, where the code value of component - dbSNP Id ({"coding":[{"system":"http://loinc.org","code":"81255-2"}]} is rs587778247.
Bonus point: More parameters can be added for searching

Scenario 5 HLA Typing

More on Scenario 5

Action: (FHIR Client) Create an HLA genotyping genomics report.
Precondition: This genomicsReport instance does not exist in service prior to action.
Success Criteria: The genomics report carrying HLA information is created correctly on server and in the desired format.
Bonus point: Extensions can be added

Scenario 6 Species Identification

More on Scenario 6

Action: (FHIR Client) Search for sequences from human (Homo sapiens).
Precondition: Sequence(s) have been created where the species is defined as human (Homo sapiens).
Success Criteria: A bundle of Sequence instances where the value of the species element is Homo sapiens.

Scenario 7 Comprehensive Pathology Report

Action: (FHIR Client) Create a comprehensive pathology report which includes genetic information for a patient. A comprehensive pathology report integrates pertinent information gathered from various methods (e.g. morphology, immunohistochemistry, flow cytometry, cytogenetics, fluorescence in situ hybridization [FISH], and molecular testing [e.g., NGS]). The genomicsReport profile has the capability to support results with simple or complex genetics observations.
Precondition: This genomics report has never been created
Success Criteria: Report created correctly on server and in the desired format.
Bonus point: Extensions can be added.

Scenario 8 Sequence quality

Action: (FHIR Client) Get the quality of the sequence under consideration. In the sequencing reads, each base is assigned with a quality score generated by the sequencer, which represents the confidence of a base call. Base quality is a critical factor for accurate variant detection in the downstream analysis.
Precondition: This sequence instance has been created
Success Criteria: Target sequence instance is returned.
Bonus point: More parameters can be added for searching

Scenario 9 PGx CDS

This scenario demonstrates an implementation of CPIC Pharmacogenomic Guidelines [2] based on FHIR, CDS Hooks [3], and genomics results formatted per our latest HL7 FHIR Clinical Genomics Reporting guide [4]

Action: In this scenario, a medication order arises in the EHR (which is also a CDS Hooks client), triggering a notification to a PGx CDS Service (which is also a CDS Hooks server). The "medication-prescribe" hook also provides the PGX CDS Service with a FHIR Medication resource that includes an RxNorm code for the ordered drug. Having been triggered by the “medication-prescribe” CDS hook, the PGx CDS Service now executes a decision support rule that [1] determines if the ordered drug has a known gene interaction; [2] determines, where there is a known drug-gene interaction, whether or not the patient has genetic test results on file; [3] determines, where there are genetic test results on file, if the patient has an interacting genotype; [4] determines, where there are not genetic test results on file, if the patient needs pre-testing. The PGx CDS Service returns a CDS hooks “suggestion card” back to the EHR, with appropriate recommendations.
Precondition: Genomic observation(s) are available for some patients.
Success Criteria: Retrieve relevant observations (variants, haplotype, genotype, etc) on a pharmacogene where they exist.
Bonus point: Correctly identify that there are no observations for the gene in question.

A swim-lane overview of the scenario is here ('GACS' is a Genomic Archiving and Communication System, used to store genomic test results):

PharmacogenomicsCDS2.png

Scenario 10: Implementing Genomics Diagnostic Reports

Goal: To learn what types of local modifications sites may require when adopting the Clinical Genomics FHIR resources and profiles, and to uncover gaps in the specification to meet emerging use cases.

Action: Each client sends a locally-customized FHIR Genomics Diagnostic Report (GDR) to a common receiving server.
Preconditions: Preconfigured FHIR server without customized resources present, GDRs created by each site.
Success Criteria: Receiving server accepts the GDR post and all data elements contained within it.
Bonus point: Successful exchange of GDRs between servers, and actionable feedback/enhancement requests.

This is scenario was designed by the ONC Sync for Genes Phase 2 pilot site participants but it is widely applicable to clinical genomics use cases. For more information, contact scenario coordinator Bob Freimuth.

Final Genomics Diagnostics Reports for Scenario 10

Utah Newborn Screening Program: File:Final S4GP2 Connectathon Utah NBS.zip
Weill Cornell Medical Center: File:Completed Cornell CaT Files.zip
NMDP: File:NMDP Original and Final Files.zip

TestScript(s)

The supporting TestScripts and corresponding fixtures have been committed to the FHIR documents Github repository at: https://github.com/FHIR/documents/tree/master/connectathons/NewOrleansJan2018/Connectathon17/Genomics


Security and Privacy Considerations