Setting the Standard for Genomics and Health-Related Data Sharing: The Global Alliance for Genomics and Health

Ethical, efficient sharing of federated genomic and health related data will transform
disease diagnosis and treatment, reveal new patterns in human biology, allow researchers to expand studies of cancer and common disease associations to less common and more detailed analyses, strengthen genomic variant interpretation in rare diseases, and enable more informed clinical decisions. However, lack of interoperability—both technical and regulatory—has resulted in data that is siloed by institution, country, and disease area, locking away their potential to contribute to research and medicine. The Global Alliance for Genomics and Health convenes the research and healthcare communities to agree on common methods for collecting, storing, transferring, accessing, and analyzing data in order to overcome these barriers, enable broad data sharing that transcends the boundaries of any single institution or country, and fulfill the human right of everyone to benefit from the advances of genomic research, as mandated by the Universal Declaration of Human Rights (Assembly 1948).

As a consequence of the continued decrease in the cost of genomic sequencing, research cohorts of hundreds of thousands of genomes have become available & millions more samples are anticipated in the coming years from both research & healthcare.However, major barriers in data sharing have hindered effective use of the data including: lack of mandates for data sharing, difficulty in submitting data for sharing, inadequate resources for ingesting & storing datasets, insufficient consent for data sharing, lack of interoperability of datasets due to disparate data models & terminologies, inconsistency in pipelines generating the data & inability to address privacy issues & provide sufficient security.In order to overcome these barriers, make the most use of these data, &fulfill the human right to benefit from scientific advances as stated in the Universal Declaration of Human Rights, the research & healthcare communities must come together to agree on common approaches. The Global Alliance for Genomics and Health (GA4GH) was launched to address this need. GA4GH Technical Work Streams, spanning Large Scale Genomics, Cloud, Data Use & Researcher Identity (DURI), Clinical & Phenotypic Data Capture, Discovery, & Genomic Knowledge Standards, have developed standards & tools that are designed to overcome technical and regulatory hurdles to international genomic data-sharing.Foundational Work Streams provide guidance to both Technical Work Streams & Driver Projects in the areas of regulatory, ethics, & data security in genomics.To date, 13 technical standards &18 policies, papers, & an over-arching data sharing framework to guide regulatory practice have been released by GA4GH & can be found on ga4gh.org. These outputs are developed by GA4GH work streams & have been identified by 23 real-world genomic data initiatives—or GA4GH Driver Projects—as immediately relevant to their community/able to support their existing Driver Project efforts. Driver Projects come from nearly 100 countries & include both national initiatives like Genomics England, GEnome Medical alliance Japan &the US All of Us Research Program as well as research programs like the Human Cell Atlas &EpiShare & programs that support the clinical interface such as ClinGen &the Variant Interpretation for Cancer Consortium.These initiatives advocate, mandate, implement, & use the policies and standards in their local contexts.
There are now several examples of research & clinical care being enabled by GA4GH standards, & these numbers continue to increase. Genomics England is using the htsget API to serve all of its data from the 100,000 Genomes Program &the Genomic Medicine Service. Access to large-scale cohorts can now be facilitated by GA4GH-developed Researcher Identity &Data Use standards by automatically matching the researcher’s characteristics to the data access policy.The All of Us Research Program intends to use these standards in conjunction with the NIH eRA Commons to manage data access. Using the GA4GH Workflow Execution Service API, researchers can then execute the analysis method(s) they choose on federated datasets that may not be shareable across international boundaries.This effort can happen at scale, independent of the details of network topology & data location. In summary, a full understanding of the basis of human health & disease will require enormous datasets & global collaboration & data sharing. The GA4GH is enabling this community effort on an international scale.