GA4GH Launches ELIXIR Partnership, Diverse Driver Projects

By Allison Proffitt 

May 9, 2019 | In February, the 20th European nation signed the European Declaration, an effort to sequence and provide transnational access to at least one million human genomes by 2022—putting to rest any lingering doubts about the breadth of our data sharing challenges and the importance of international standards.

The Global Alliance for Genomics and Health is at the forefront of the work.  

GA4GH has been a cross-country collaboration from its genesis. In the first days of the Global Alliance funding came from the Broad Institute in the U.S., the Ontario Institute for Cancer Research in Canada, and the Wellcome Trust in the U.K. Early Driver Projects include Genomics England, Australian Genomics, and the All of Us research program. Now Global Alliance projects bring together resources from 97 countries around the world.

“It's really important that we represent the entire global community in this project,” says Heidi Rehm, one of two Vice Chairs for the Global Alliance, and an institute member at the Broad Institute of MIT and Harvard. “In genomics there is such a huge advantage of data sharing from every country, every race, and ethnicity.

Last week, the Global Alliance announced a strategic partnership with ELIXIR, a group uniting Europe’s leading life science organizations in managing and safeguarding the increasing volume of data being generated in publicly funded research. The two groups are extending their long-term working relationship to facilitate the responsible sharing of sensitive genomic data across Europe. The effort will enable the creation of a secure infrastructure to store, access, and analyze genomic and sensitive health data, and create virtual cohorts with tens of millions of participants. “International collaboration is needed to provide recognized, secure, standardized, documented, and interoperable services,” write the two groups in the Strategic Partnership agreement.

ELIXIR has a long history with the Global Alliance, and the strategic partnership represents a graduation, of sorts, in the relationship. ELIXIR joined GA4GH in ​2014 ​and ​has ​formally ​contributed ​to ​GA4GH ​projects ​since ​2015. ELIXIR’s Beacon project was one of the first GA4GH Driver Projects.

Driver Projects are formal collaborations first announced in 2017 to help identify, develop, and pilot data sharing frameworks and standards in real world settings. Now, nearly two years later, the Driver Project approach has been, “an incredible success,” says Kathryn North, Co-Vice Chair for the Global Alliance and Director of Australian Genomics. “It's really accelerated our ability to develop really relevant and useful standards, and to immediately apply them, because we've engaged very strongly internationally.”

In February, the Global Alliance for Genomics and Health (GA4GH) named seven new Driver Projects for 2019 including ​Human Heredity and Health in Africa (H3Africa), Genome Medical Alliance Japan (GEM Japan), ​European Joint Programme on Rare Diseases (EJP RD), ​Swiss Personalised Health Network ​(SPHN), EUCANCan, EpiShare​, and​ Autism Sharing Initiative​.

North is most proud of the international reach of the new class of projects. The logistics of sharing clinical and genomic data can be a challenge whether you’re sharing across the hall or across the globe, but North believes, “the most efficient and effective way to look at how we can implement these standards widely is to look at it on a country to country basis.”

Country-level genome projects—ones with government investment—are particularly important, North argues. “They've really seen the power of joining in at a country level and it's always a two-way exchange,” North says. She hopes that the Global Alliance can pave the way for other governments to make country-wide investments as well. A country that wants to set up a new national genomics program can immediately adopt standards and tools that have been vetted elsewhere and plug them in. “I think it's really facilitating and speeding up the ability to do genomics at scale in clinical practice on a country scale,” she says. “We're really starting to dig deep to solve some of the problems about sharing across national borders. Although, there's still a lot to do.”  

Genomics England has led the way on national genomics programs, North says, but she pointed out that each country program involved in the Genomics Alliance has had a slightly different focus, and so has contributed new data. “The All of Us program is looking at how you take it out in the communities with huge biospecimen banks. H3Africa is looking from a prospective of infectious diseases and complex diseases and genetic risk factors. And the Swiss initiative is certainly initially very focused on rare disease,” she says. “Whereas in Australia our initiative is actually starting with a patient in the clinic and working backwards from there. So you can sort of see that those different perspectives end up being very complimentary, especially when we get a couple of years down the line.”

Diverse Inputs Needed

In the latest call for Driver Projects, the Global Alliance emphasized diversity as one of its major criteria—diversity in race, gender, and indication or disease area of focus. Rehm believes that diversity within even the groups building standards is vitally important. 

“When you're developing a standard, if one group develops a standard for their one use case, well, it's going to work great for their one use case. But then you send it out to the communities, and the next group tries to take it on and says, ‘Well, this doesn't work for me because my environment is very different than yours.’ So we have many factors that influence how well the standards we build work.”

The effect will be amplified as we move toward genomics in clinical care.

“The vast majority of genomes in the next five to ten years are going to be generated in clinical practice,” North says. “We're talking about how you negotiate the different health systems with different ways of capturing data as electronic medical records, and at different stages of maturity.”

Rehm agrees. “If you think about the infrastructure in developing countries, and other communities that really have very different social and political systems, then the standards that will work in their systems sometimes are a little different,” she says. “Undoubtedly, if you ensure interoperability and development of those standards with the global community, the likelihood that data will come from those communities is much higher.”

North calls the Global Alliance a coalition of the willing, and says the vision is catching on. “I think there's this little groundswell of seeing the value of this. A lot of the data sharing happens in a pre-competitive space anyway. We also really want to make sure that the standards that are developed are all freely available. A lot of work is going into just making sure that these are launched and available, because that's the most rapid way we're getting to promote data sharing.”