About us
Learn how GA4GH helps expand responsible genomic data use to benefit human health.
April Connect 2023
19 Apr 2023
GA4GH Connect 2023 was held 19 to 21 April 2023 at the Royal Society of London, UK, and virtually. Read more below.
The working meeting gathered 188 in-person attendees and 219 virtual attendees, representing more than 30 countries. Attendees joined 26 working sessions over three days. Watch recordings and read minutes of the sessions. GA4GH celebrated its 10th anniversary in a fireside chat with Research Data Alliance (RDA) and a reception with a toast from GA4GH CEO Peter Goodhand.
19 April
Opening remarks
The meeting was initiated by Chair Ewan Birney, who outlined the expected behaviour and . a summary of the agenda for the meeting.
Fireside chat: reflections and insights from a decade of GA4GH and RDA
GA4GH CEO Peter Goodhand and Research Data Alliance (RDA) CEO Hilary Hanahoe met for a fireside chat, where they reflected on both organisations’ tenth anniversaries, discussed major milestones achieved, and shared future plans towards advancing open and FAIR data sharing.
Work Stream updates
Representatives from GA4GH Work Streams and initiatives presented updates on their progress, goals for the Connect meeting, and next steps for implementation.
GA4GH implementations
Speakers from Australian Genomics and the European Genomic Data Infrastructure highlighted notable practical applications of GA4GH products in the real world.
VA specification status and implementation updates
The Variant Annotation (VA) specification will provide a framework to define schema for different types of variant knowledge statements that are rooted in a common underlying conceptual model so that they can be understood and used together in working data systems.
To date, the VA team has focused on an implementation-driven development approach, wherein Driver Projects directly define initial drafts of statement models to support working systems with real data. For example, ClinGen and the VICC are developing models for Variant Pathogenicity and Therapeutic Response Statements, to support the exchange of ClinVar and CIViC data between their variant curation platforms. As these implementation models are tested and refined in real-world applications, the VA team will oversee their coalescence and hardening into standard models that can be used by the broader community.
This Connect session provided an overview of an initial v1.0 release of the VA spec and gave updates about work by the VA team. They discussed specific challenges related to coordination with other GKS standards, as well as infrastructure needed to support an implementation-led development paradigm. Finally, they engaged the larger GA4GH community for feedback and requirements.
The key takeaways from the session are:
- there’s a need for common modelling framework (LinkML was discussed, but several in room recognized need for high-level agreement);
- VA vision for core information model informing development through standard profiles was identified.
REWS General Meeting
The REWS General Meeting brought together all contributors across all subgroups to discuss topics relevant to the full Work Stream. This session included subgroup updates, updates from Driver Projects, community updates, and discussion around the road map directions.
Federated compute
This session showcased the status of various GA4GH-based compute/cloud solutions through live demos and presentations, with a focus on compute federation. Presentations were followed by an open-floor discussion about the current limitations and specific requirements of federated compute, and their implications for future developments of relevant GA4GH standards.
The key takeaways are;
- interop on WES is increasing, e.g. cwl-WES x WESkit x Sapporo;
- interop on TES is increasing, TESK x Funnel;
- blockers are mostly on interoperable IAM/AAI.
Diversity in Datasets
This session presented the policy developed by the Diversity in Datasets subgroup of REWS. Participants workshopped ways to strengthen this policy proposal and ensure the output can be useful and accessible in practice.
Federated variant matching
This cross-Work-Stream meeting supported efforts to develop a federal variant matching platform across different databases.
Collaborative evaluation of “Future of VCF” resource benchmarking results
The goal of this session was to discuss together the results of “Future of VCF” resources/initiatives benchmarking according to benchmarking metrics drafted by the working group. They evaluated and aggregated the results to produce a summary document highlighting trends, conclusions, and next steps.
The key takeaways from the session are:
- there are several steps required to get the next VCF spec released, including outside review;
- work on the next VCF spec release has been hindered by lack of maintainers;
- GA4GH blogs are a useful format for sharing what they’ve learned so far in this working group prior to being able to do a formal release of the next VCF specification.
DRS and Data Connect
This session explored how standards from the Cloud and Discovery Work Streams can be used together to:
- address the need to obtain additional data about a DRS object;
- revisit how Data Connect handles the need for bundles in DRS;
- identify actions for DRS and Data Connect.
The session also aimed to bring closure through a path forward involving both Work Streams.
The key takeaways from the session are:
- best practices for working with metadata are important;
- the DRS 1.3.0 develop branch frames they need for compound object support.
20 April
Crypt4GH: developments and demo
The speakers showed developments happening with Crypt4GH and demonstrated existing use cases.
They key takeaways are:
- make existing cloud APIs provide and understand data in crypt4gh format;
- the format needs to be updated;
- understand how the format is being used in the EGA and how beneficial it is.
Consent for the sharing of biological materials and data in genetics/genomics research
This session examined the dynamic role of consent for the processing of genetic/genomic biological materials and associated data in an evolving European data-driven society. It explored how the evolution of consent for health-related research — with specific attention to genetic/genomic biological materials and associated data — is increasingly driven by data and AI policies in European legislation from the GDPR to the DGA, EHDS, AI Act, and EOSC.
The session was interactive and created a space for open discussion on a critical topic of importance to ethics and citizen rights, at the interface between data, AI systems, and genetic/genomic research.
The session elucidated the developing legal and ethical relationship between the data subject and his/her data, examining how evolution in consent frameworks and requirements impacts fundamental concepts of privacy, control, and governance in genetic/genomic research.
The session looked at the ongoing modifications of consent, from “taking control over one’s data” to questions of “legacy consent” to “broad consent” to “informed choice” to “data donation” to “a common European data altruism consent form” for genetic/genomic research. Where will consent land for human biological materials and associated data in genetic/genomic research in the evolving EU context, where machine learning and AI are developed in regulatory sandboxes for uses in the European Health Data Space?
The session confronted genetic/genomic research with the evolving role and understanding of consent within the fundamental EU Charter rights, and the ethics of high-risk AI systems developed for data management in data sharing and open-science structures.
GKS maturity model — versioning strikes back (part 1)
Development of specifications to address new concerns in genomic knowledge sharing requires iterative improvement over time. When these components are designed to work interoperably with other data standards and frameworks, there needs to be clear communication with the GA4GH user community about the maturity and availability of the data classes from these specifications.
The GKS Work Stream has developed a maturity model and versioning system for communicating the readiness and stability of new specifications. GKS has applied this in a variety of scenarios from experiences in the development of the VRS and VA specifications.
Technical Alignment Subcommittee (TASC) meeting
Pagination is currently handled by several GA4GH specs, but there is not necessarily interoperability between them. A goal for this session was to determine whether there is a benefit in converging upon a minimal common pagination schema that existing specs can make use of, and discuss what that would look like. The session will showcased the group’s work on the TASC governance document.
Public Attitudes for Genomic Policy
Speakers presented and reviewed the current work that has been produced by the Public Attitudes for Genomic Policy group to date. They then led a workshop to identify what is missing from our knowledge of public attitudes towards genomics and genomic data sharing to help us inform a further public attitudes study.
GKS maturity model — versioning strikes back (part 2)
Development of specifications to address new concerns in genomic knowledge sharing requires iterative improvement over time. When these components are designed to work interoperably with other data standards and frameworks, there needs to be clear communication with the GA4GH user community about the maturity and availability of the data classes from these specifications.
The GKS Work Stream has developed a maturity model and versioning system for communicating the readiness and stability of new specifications. GKS has applied this in a variety of scenarios from experiences in the development of the VRS and VA specifications.
Equity, diversity, and inclusion (EDI) workshop — strategic planning
In this session, GA4GH staff (with support from the EDI Advisory Group) presented their plans for an EDI Strategic Plan. This plan is being developed as a tool to hold GA4GH accountable for real and actionable steps toward creating a more equitable, diverse, and inclusive environment. The session was used to collect feedback from the GA4GH community to ensure that the plan’s priorities and goals reflect their experiences with issues of EDI at GA4GH.
VRS 2.0 — a new hope
The GA4GH Variation Representation Specification (VRS) was initially released in 2019. Since then, GKS has released two minor version updates, and an upcoming v1.3 release supporting genotypes. With the growing uptake of VRS in knowledge-base and genomic information systems, as well as within larger message structures (e.g. Beacon and Phenopackets), the group has identified several key features for enabling use of VRS in these settings.
In this session, they reviewed these key features and outlined the vision for the second major version of VRS.
The recording will be valuable for anyone interested in variation representation integration into system-to-system exchange and genomic databases.
AAI/Passports update
This session featured updates on what’s new in the v1.2 Passports release and hosted an open discussion about possible next steps for AAI and Passports.
Beacon, Phenopackets, cohorts — combining standards for discovery and exchange
This session brought together the Phenopackets and Beacon development teams, as well as those interested in cohort discovery and querying, to explore how to optimise Phenopackets and Beacon integration and enable Phenopackets queries across cohorts.
21 April
CINECA use cases and GA4GH standards: lessons learned and gaps identified
This session demonstrated how GA4GH standards have been used within the CINECA project, and detailed the lessons learned in trying to implement a set of discovery, access, and analysis use cases.
The group identified the gaps and issues that exist, ways existing GA4GH standards can be further developed or new standards defined to can address these issues, and which Work Streams may have an interest in addressing these issues. They also identified other projects, GIF projects, and Driver Projects able to address these issues in the future to ensure the lessons learned in CINECA are not lost once the project ends.
Cancer Community
This session was a meeting of the GA4GH Cancer Community (G-CC). G-CC was established in 2020 as a forum for researchers, clinicians, and technical experts to define, discuss, and advance real-world cancer use cases through the implementation of GA4GH interoperability standards. In this session, participants learned more about developments in the cancer genomics field from an invited speaker. The discussion focussed on ways that GA4GH products could be used to advance real-world oncological use cases.
Applying GKS in EHR
An overview of the efforts to ingest genomic knowledge across resources and use genomic knowledge standards (GKS) to match genomic findings from the clinical setting to these aggregate knowledge sources was presented. The group also reviewed how genetic testing orders, results, and associated genomic data are transmitted between labs and providers using Epic.
Gaps in the representation and completeness of genomic data, and how this is associated with genomic knowledge sources, was reviewed, and challenges presented by these gaps were assessed.
The session concluded with a discussion of how development and adoption of GA4GH genomic knowledge standards can help address these challenges and reduce burden on genomic data providers and consumers.
The key takeaways from the session are:
- Epic supports a spectrum of users and workflows with respect to genomic data and knowledge, from interpretation to patient-facing clinical decision-making;
- recommend to Epic to focus efforts on representing variation and matching variation to knowledge;
- GKS have not solved issues related to test methodology and what can be said when variation isn’t specifically reported.
GKS — return of the road map
This session was a recap of other sessions relevant to GKS activities at April Connect 2023, with cross-team planning for the 2023 Q2/Q3 road map.
Anyone with an interest in developing and applying the GA4GH genomic knowledge standards should watch the recording.
The key takeaways from the session are:
- formally defining VA modelling process was based on experience to date with VICC/ClinGen effort;
- maturity model seemed like good start to put into practice;
- develop a SIG for Categorical Variation;
- integration of non-inherent properties with VRS v2.0 and cross-domain modelling;
- SA and VR cross-activity would be a model feature;
- develop a Gene-Disease profile for VA.
Experiments Metadata
Metadata describing -omics experiments is crucial to make sense of results that are available to the community. For instance, a CRAM document provides results of a sequencing experiment, but one needs to understand whether the data is from a whole genome sequencing experiment, RNA-Seq, ATAC-Seq, etc.
In this GA4GH Study Group, they assessed whether a minimal specification could be established to better characterise -omics datasets.
The goal was not to organise full recipes to make an experiment reproducible, but rather to provide enough information to understand what the experiment is about. What “enough information” means is subject to further discussion in this Study Group.
DRS test-a-thon
System builders were invited to test or pre-test their implementations. They were also welcome to bring feedback for aggregation and comparison with other implementations, leading to proposed changes to the test suite and specifications.
The key takeaways from this session are:
- for the DRS group in the Cloud Work Stream…
- what about extra fields in service-info? Is that allowed? Should it be discussed at the next GA4GH Cloud Work Stream DRS meeting. What does the TASC team think?
- for the test suite devs…
- can they add “checksum” to the drs_object_info so the compliance test suite can check the file downloaded?
- can they add “negative” tests to the test suite? E.g. “expectation: ‘404”’ for a drs ID that requires auth but auth is null.” In that way they can test that the server is doing what’s expected.
Sequence Collections: how to uniquely identify and compare genomes
This session described the current status of the Sequence Collections project and connected interested individuals with relevant work.
The key takeaways from this session are 3 discussion topics:
- can they return the collection in an alternative format? (Answer: Maybe in the next version; for now they don’t want to mess with it.)
- sequence type, and alphabet…
- Should one or both of these be specified in the spec?
- Should they be inherent?
- how should they reserve attribute names? Should they use namedspaced attributes? Can this be uniform across other GA4GH standards that also have possible extension?
Discovery road-mapping
The session provided discussion and interaction to identify a model which reflects the complementary roles of the Discovery standards.
They also discussed findings from the following April Connect sessions:
- Beacon, Phenopackets, cohorts — combining standards for discovery and exchange
- Experiments Metadata
- DRS and Data Connect
Ethical preparedness for future pandemics and epidemics
The GA4GH Infectious Disease Community convenes international groups focused on advancing human genomics for the purpose of treating and diagnosing infectious diseases. It interfaces and collaborates with the pathogen genomics community, specifically through a partnership with the Public Health Alliance for Genomic Epidemiology (PHA4GE). In this session, Anja Bedeker (PHA4GE) facilitated a discussion on research ethics, including benefit sharing and risk analysis as it relates to infectious disease.
GA4GH Implementation Forum (GIF) and close of meeting
The GA4GH Implementation Forum (GIF) was officially launched at this session! GIF is where the potential and power of GA4GH standards are recognized. The forum guides diverse scientific and clinical groups as they identify real-world use cases that can benefit from the end-to-end implementation of GA4GH products.
In this session, participants learned about:
- the structure of GIF;
- the four GIF Projects that will be going through the process first;
- how to participate in these GIF Projects.
There was also time at the end of the session for participants to propose their own GIF Project ideas.
Categories
