Genomic medicine has the potential to transform healthcare by integrating patients’ genomic information into clinical care1. It offers the possibility of a previously unprecedented level of precision in diagnosis and treatment of patients, particularly in oncology and rare diseases2,3,4,5. The genomDE initiative spearheads the idea of increasing the use of genomic analysis in patient care. It aims to establish a nationwide data infrastructure that bridges the gap between healthcare provision and research.6 The goal is to enhance the understanding of diseases and their subtypes at both clinical and pathophysiological levels. This integrated approach may facilitate more comprehensive analyses and foster deeper diagnostic capabilities and personalized therapeutic discovery6,7. Cross-entity feasibility within the genomDE infrastructure is of importance, with data provided by the German Familial Breast and Ovarian Cancer Consortium and other participating networks.

As genomDE connects specialized centers across Germany, aiming to create a nationwide platform for genomic medicine, institutions such as German Human Genome Archive (GHGA) play a central role in archiving and managing genomic data securely. The GHGA infrastructure is a federated network with GHGA data hubs located at seven sites throughout Germany. Thus, interoperability of said data is of high importance. The primary users include clinicians, researchers, and regulatory authorities. The system is designed to allow controlled access, ensuring that researchers can use anonymized or pseudonymized data while maintaining patient privacy and security.

genomDE not only focuses on its national strategy but also sets ground for international exchange of genomic data and therefore participates in the international consortia 1+ Million Genomes Initiative by the European Union. This initiative aims to provide a federated infrastructure for storage and analysis of genomic data. The federated approach, the sensitive nature of clinical data and the variety of genomic data format make the integration of interoperability a necessity. This is why the genomDE project prioritizes the concept of interoperability—the seamless exchange and functional use of information across diverse information technology (IT) systems. Furthermore, genomDE lays eyes on previously established interoperable data models and data sets to be able to exchange information nationally and internationally. Interoperability, however, is multifaceted, encompassing organizational, technical, semantic, and syntactic dimensions7. This feasibility study focuses on two key aspects: semantic and syntactic interoperability. Semantic interoperability is achieved through the adoption of international terminologies and ontologies, ensuring that the meaning of exchanged information is universally understood. For instance, the Sequence Variant Nomenclature (HGVS) provides a structured system for naming genetic variants at the DNA, RNA, and protein levels8. And Systematized Nomenclature of Medicine – Clinical Terms (SNOMED CT) is the largest medical terminology comprising more than 350.000 concepts to represent medical data. Syntactic interoperability pertains to the use of uniform data formats, facilitated by IT standards that dictate the structure of data.

The Health Level Seven (HL7) organization’s “Fast Healthcare Interoperability Resources” (FHIR®) standard was developed for healthcare purposes and is also gaining popularity in the domain of health research9. Previous work has explored the integration of genomics and clinical data using FHIR ®, such as vcf2fhir for genomic data conversion and SMART on FHIR Genomics for standardized applications, highlight the ongoing advancements in FHIR-based clinico-genomic integration10,11,12,13,14,15,16,17,18,19.

FHIR consolidates functionalities of its predecessors to offer a robust framework for data exchange, utilizing web technologies to enable comprehensive system integration. Within FHIR, resources serve as the building blocks, defining data elements, cardinalities, value sets, coding systems, and inter-resource references, which, for example, can link medication administration to a specific healthcare provider20. In addition, FHIR has been used for several use cases within genomic medicine: The National Institute of Health Cloud Platform Interoperability (NCPI) program is creating a federated ecosystem for genomic data that enhances researchers’ access to all data types21. The program’s external partners comprise three working groups addressing interoperability challenges specifically, one of them being the FHIR working group22. NCPI’s FHIR IG contains high level information and therefore was not included in our precise mappings. Furthermore, the Genomics Reporting IG by HL7’s Clinical Genomics Work Group23 and the GA4GH Phenopackets Schema represents significant advancements in standardizing the representation and exchange of genomic and phenotypic data24. Within the Genomics Reporting IG Version 2.0.0 fourteen FHIR profiles were developed enabling the representation of known and de novo variants, of simple and complex nature, somatic or germline origin23. It should be noted that a new version of the Genomics Reporting IG is now available that was not published yet at the time of the mapping process.

The specification also allows for deriving implications based on observed genomic characteristics regarding disease pathology, medication recommendations, diagnostics, and suitability for transplantation. The Phenopacket Schema ISO 4454:2022 defines a computable representation of clinical data to enhance the application of reusable analysis pipelines. There is a draft version of a FHIR representation of the Phenopacket Schema with an associated IG published on HL7 FHIR. As the Phenopacket provides a structured representation of phenotypic and genomic data that current FHIR Genomics standards do not yet fully cover and given its adoption by GA4GH and ongoing efforts to align it with FHIR, we anticipate that it will play a key role in future interoperability standards. Within the GA4GH Phenopacket Schema, a single person or bio sample is characterized in a Phenopacket and linked, among other things, with detailed phenotypic descriptions, genetic information, diagnoses, and medical treatments. A PhenotypicFeature represents the central element of the Phenopacket schema. Phenotypic characteristics such as symptoms, laboratory results, histopathological and radiological findings can be represented in a PhenotypicFeature with modification and qualification concepts24,25.

On a national level, the MII, formed by representatives from German university hospitals, research institutions, and the private sector have been developing the MII core dataset (KDS) based on FHIR and international terminologies such as SNOMED CT. The KDS is divided into basic and extension modules26. The extension modules represent specific medical disciplines such as molecular genetics, also called “MolGen Befund”, or in English: Molecular Genetic Findings Report27. The Extension module “MolGen Befund” or the Molecular Genetic Findings Report focuses on the structured representation of genetic characteristics with a FHIR IG28. genomDE played a key role in mapping and aligning the Molecular Genetic Findings Report dataset within the broader MII framework.

This study aims to assess the feasibility of mapping Phenotype sample data defined by genomDE to the HL7 FHIR framework, considering the foundational work of the already established international Genomic Reporting IG, GA4GH Phenopackets, and the national MII’s KDS. By examining and validating FHIR test data and query searches derived from molecular genetic findings, the study endeavors to enhance the precision and efficiency of international data exchange within genomic medicine, while considering national goals and requirements.