Poster Number
23
Poster Title
Enabling the Clinical Integration of GA4GH VRS through Interoperability with HL7 FHIR
Authors
Salem Bajjali, M.S., Aly Khalifa, Ph.D., Xianfeng Chen, Ph.D., Sarah R. Senum, M.S., Robert R. Freimuth, Ph.D.; Mayo Clinic
Abstract
As genomic testing expands, it is becoming increasingly critical to bridge clinical systems and public knowledge bases. The GA4GH Variation Representation Specification (VRS) supports structured, computable, and semantically robust representations of genomic variation, which are essential for systems to exchange genomic data effectively. Several pilot projects have successfully implemented VRS to represent publicly available genomic data sets. To enable widespread adoption among clinical software systems, it is imperative to integrate VRS into clinical data standards.
HL7 FHIR® is a clinical messaging standard that supports electronic health record systems (EHRs). In some countries, clinical systems are required to adopt the FHIR standard. Therefore, in order to lower barriers to interoperability among systems that support VRS and systems that support FHIR, the two specifications must be aligned so that they can be used to seamlessly exchange data within infrastructure that supports workflows that span a wide spectrum of clinical and research use cases.
To advance interoperability, we are incorporating VRS principles into new FHIR standards. We have performed semantic (conceptual) mappings between the specifications and implemented bidirectional technical transformations in Python and Jupyter notebooks. We also developed an interactive Github Codespace environment that allows users to explore translations between FHIR and VRS data structures. Evaluation of the implementation using example data curated by domain experts identified areas of alignment, as well as opportunities for further harmonization. This effort is ongoing and aims to achieve improved interoperability with continued convergence of the specifications.
This work demonstrates the successful embedding of VRS principles within the FHIR specification, and it represents substantial progress towards enhancing genomic data interoperability between HL7 FHIR and GA4GH VRS – an important step in enabling the exchange of genomic data to improve health.
HL7 FHIR® is a clinical messaging standard that supports electronic health record systems (EHRs). In some countries, clinical systems are required to adopt the FHIR standard. Therefore, in order to lower barriers to interoperability among systems that support VRS and systems that support FHIR, the two specifications must be aligned so that they can be used to seamlessly exchange data within infrastructure that supports workflows that span a wide spectrum of clinical and research use cases.
To advance interoperability, we are incorporating VRS principles into new FHIR standards. We have performed semantic (conceptual) mappings between the specifications and implemented bidirectional technical transformations in Python and Jupyter notebooks. We also developed an interactive Github Codespace environment that allows users to explore translations between FHIR and VRS data structures. Evaluation of the implementation using example data curated by domain experts identified areas of alignment, as well as opportunities for further harmonization. This effort is ongoing and aims to achieve improved interoperability with continued convergence of the specifications.
This work demonstrates the successful embedding of VRS principles within the FHIR specification, and it represents substantial progress towards enhancing genomic data interoperability between HL7 FHIR and GA4GH VRS – an important step in enabling the exchange of genomic data to improve health.