CG WG Call Notes leading to 2014 January WGM

Proposed Agenda

• San Antonio WGM draft agenda
• The Clinical Genomics Statement DIM
  • The label Genotype?
  • The association to Family History
  • The bidirectional association of phenotype and genotype

Attendees

• Amnon Shabo,Ruby Russell, Siew Lam, Scott Bolte

Draft Minutes

• Briefly discussed the draft agenda
• Decided to defer the DIM discussion to next week, to let people the chance to go through the latest draft model and the email correspondence around it, so that we could discuss it next week with more feedabck

Proposed Agenda

• Continue the CG DIM discussion

Attendees

• Amnon Shabo,Larry Babb, Ruby Russell, Mollie, Siew Lam, Scott Bolte, Grant Wood

Draft Minutes

• Amnon walked through the second draft of a "Clinical Genomics Statement" model
• Mollie suggested to add a reference from the observed phenotype class to the family health history
• Mollie suggested to add a reference from the observed phenotype class to relevant anatomic pathology reports
• Discussed the "genotype" term and it was suggested to go with "Variant"
• An open issue is whether to designate the variant as a DNA variant or genetic variant or just a generic variant
• An open issue is the whether we need to have the genotype-phenotype association bi-directional so that a phenotype could be associated with multiple genotypes
• If we stick with the unidirectional association as it is in the current draft, then it's possible to instantiate multiple instances of "clinical genomics statement"
• Yan asked asked about the direction and multiplicity to the indication class
• Lam asked about multiple genotypes that can be inferred by one of them
• Larry asked whether a number of phenotypes can be grouped by some category; Amnon suggested that the Phenotype class could be a UML package where such grouping and other structured could be made possible

Proposed Agenda

• Continue the CG DIM discussion

Attendees

• Amnon Shabo,Larry Babb, Grant Wood, Ruby Russell

Draft Minutes

• Amnon walked through the first draft of a "Clinical Genomics Statement" model, that could be a core component of the CG DIM and be used in a document, message, resource, etc.
• The is a conceptual model and at this point it serves as a source of discussion around the main constructs we would like to see present when describing genotype-phenotype associations, for example:
  • Explicit representation of the association between a genotype to a number of phenotypes
  • The distinction between interpretive phenotype and observed
  • The separation of the interpretation as a distinct observation with its attributes such as time, method, performer, etc.
  • Links to knowledge sources used for the interpretation
  • Machinery for encapsulation of key raw omics data within a clinical genomics statement, as well as links to the full-blown omics data (e.g., NGS sequencing data)
  • Optional indications representing the triggers for performing genetic testing or making genomic observations
  • Specimen and genomic source class

Proposed Agenda

• Present and discuss in detail the FHIR resources

Attendees

• Amnon Shabo,Ruby Russell, Grant Wood, Gil Alterovitz, Larry Babb, Tom (CBMI), Siew Lam, Scott Bolte

Draft Minutes

• Amnon's comments:
  • How do the FHIR resources represent genotype-phenotype associations?
  • How do you distinguish between interpretive phenotype and observed phenotype?
  • How do we characterize a genotype-phenotype association while keeping each side of the association modular, e.g., representing observed phenotype using other standards for clinical data?
  • Interpretation attributes should be separated from the class representing the genomic observation, so that it can have its own time stamp, method, performer, etc.
  • Should distinguish between types (levels) of analyses mentioned in the various FHIR resources, e.g., Sequencing Analysis is different than the Genetic Analysis - the former is lower level analysis at the genomic side, while the latter is more of a clinical genomics analysis
  • Having FHIR Resources named after common representations like VCF and GVF is not flexible and should be abstracted a bit, e.g., through a wrapper of VCF-like formats that could include various existing formats and new ones in the future
• Larry:
  • the Genetic Analysis FHIR Resource roughly follows an older version of the v2 message for genetic testing results, and that version has been recognized to be limited and not being able to accomodate all use cases
  • e.g., genotypes and phenotypes could have many-to-many relationships, and not necessarily as defined by some testing order
• Amnon:
  • this discussion is held in the context of the effort to develop an agreed-upon Clinical Genomics Domain Information Model(s)
  • therefore, all lessons learnt from existing specifications and pilots should inform the design of the CG DIM in a way that will result in an all-encompassing model that can be derived to produce different standards but still aligned semantically
  • so, it's a top-down approached, informed by bottom-up reality...

Proposed Agenda

• Discuss and vote on whether to endorse the familial relation codes harmonization proposal
• Sorting out our projects in Project Insight

Attendees

• Amnon Shabo,Ruby Russell, Grant Wood, Gil Alterovitz, Mollie, Larry Babb, Yan Heras, Lloyd McKenzie

Draft Minutes

• Motion to endorse the familial relation codes harmonization proposal passed unanimously
• Discussion of projects listed under CG in HL7 Project Insight:
  • HL7 CG-omics Domain Analysis Model
    • send email to group announcing the intention to to retire this project
    • if content is informative for other work - please advise group and lead discussion on a weekly call
  • Gene Expression (project 867)
    • send email to group to retire this project,
    • if content is informative for other work please advise group and lead discussion on a weekly call
  • Clinical Sequencing
    • rename and expand scope, after January ballot
  • Genetic Variation CMET (project 693)
    • deprecate
  • Pedigree Topic - ISO Ballot (project 479)
    • put on hold/inactive -
  • Cytogenetics Lab Result (project 663)
    • status should equal published and complete but doesn't - need to clarify
  • Canonical Pedigree - Family History (project 729)
    • should list Published and completed because that's where we published the US Implemenetaion Guide for the Pedigree spec - need to clarify
  • Pedigree release 2
    • on hold
  • Aligning the CG DAM, DIM, and FHIR Resources is easier as these are all under development
    • goal to make first step in Jan ballot

Proposed Agenda

• Discuss the familial relation codes harmonization proposal

Attendees

• Amnon Shabo, Siew Lam, Ruby Russell, Grant Wood, Scott Bolte, Lloyd McKenzie, Phil Pochon, Chris Pack, Mollie, Larry Babb, Yan Heras

Draft Minutes

• Discuss the familial relation codes harmonization proposal
• Mollie: needs to run due diligence and understand its implications on current standard specifications and implementations
• Amnon:
  • it's independent of the FHIR discussion and of any type of standard structure like v3, v2, CDA or FHIR
  • it's an improved vocabulary that all standards can utilize
  • standards can bind to it dynamically or extract subsets of it and bind to statically
• Motion to send out the harmonization proposal to the entire CG listserv and ask for comments, as well as reach out to Kevin Hughes team to check their feedback and possible implications on their implementation of the Pedigree spec that utilizes the current version of the vocabulary
• Lloyd seconded it
• Lam will send out an article with recommendations from genetic counselors
• Mollie will look into HGVS recommendations
• No objections or abstains and motion passed

Proposed Agenda

• Update on the Global Alliance by Gil

Attendees

• Amnon Shabo, Siew Lam, Gil Alterovitz, Ruby Russell, Grant Wood, Scott Bolte

Draft Minutes

• Discussed the correspondence of Scott, Lloyd and Kevin on Family History and FHIR

Proposed Agenda

• Discuss the CG DIM project

Attendees

• Amnon Shabo, Siew Lam, Grant Wood, Gil Alterovitz, Mollie, Phil Pochon, Yan Heras

Draft Minutes

• Amnon went through a slide deck, each containing a model underlying one of the specs developed by the HL7 Clinical Genomics group, and then presented a sketch of thought on where we could go with the DIM

Proposed Agenda

• Discuss the revised Project Scope Statement of CG Domain Information Model(s)
  • See discussion on modeling issues at the project wiki page: http://wiki.hl7.org/index.php?title=Clinical_Genomics_Domain_Information_Model(s)_Project#Modeling_Discussion

• Need to submit the Cambridge WGM Meeting Minutes by the end of this week

• Change of WGM schedule
  • From Wednesday & Thursday to Tuesday & Wednesday
  • Do we need all session in both days?
    • Accommodates joints with AP and OO with solo meetings before the joints
    • So far we had a day and a half
    • We could run from Tuesday Q2 to Wednesday Q3

Attendees

• Amnon Shabo, Siew Lam, Grant Wood, Larry Babb, Gil Alterovitz, Jenny, Ruby Russell, Mollie

Draft Minutes

• Grant will start the Cambridge WGM minutes by inserting the attendees in each session; Amnon will add to it minutes as he can recall and send out to the mailing list so that people can add to it
• Discussed the revisions to the CG DIM PSS and hearing no objections decided to resubmit to DESD
• Discussed initial modeling considerations of the DIM
• Amnon brought up the GTR, v2, v3 and FHIR models and made comments on the current differences
• Is research use case part of the DIM?
• Mollie said it should be clinically oriented with transnational capabilities at the most
• Amnon said it's Universal in the sense that we start from common denominator and then allow each environment to constrain it as need; e.g., for clinical it should be highly constrained