On March 4th, over 200 experts in genomics, healthcare, biomedical research, bioinformatics, ethics, and patient advocacy converged on London. Their destination: The Wellcome Trust at 215 Euston Rd, host of the first meetup of the Global Alliance for Genomics and Health. As a partner of the alliance, we were thrilled to join the inaugural meeting. Here’s a short recap of the alliance’s mission and impressions from the meeting.
What is the Global Alliance and where did it come from?
As genomic data is being generated at a mind-boggling pace, realizing its potential hinges on researchers and clinicians accessing and interpreting this data in a unified way. The goal of the alliance is no less than to enable the “sharing of genomic and clinical data in an effective, responsible, and interpretable manner” for the benefit of human health – an ambitious aim that implicates issues ranging from data security and data privacy to ethics and computational efficiency.
The alliance first made a public appearance in June 2013 as the “global alliance” in a draft white paper that was widely circulated and signed by dozens of attention-raising names in genomics. This original manifesto makes salient points about the advantages of cloud computing that we could hardly have said better ourselves.
The current mission
Since June 2013, the alliance has not only evolved a more focused name, but has also refined its mission and goals. Four working groups have been established to tackle the major hurdles of genomic data sharing:
- The Regulatory and Ethics Working Group works on guidelines and ethical frameworks for global sharing of genomic and clinical data.
- The Genomic Data Working Group develops data formats and APIs for the representation, storage, and analysis of genomic data
- The Security Working Group focuses on the technological aspects of data security, user access control, and audit functions
- The Clinical Working Group is working to unify existing ontologies that are used to link genomic data with phenotype and clinical data
Divergent as the work areas of the groups may be, their representative speakers delivered the coherent message at last week’s meeting that, regardless of technical focus, all policies, security standards, and algorithms developed will, ultimately, be measured only by how effectively they support one goal: improving healthcare.
The Inaugural Meeting: off to a promising start
The meetup featured presentations from the four working groups as well as existing data sharing initiatives from around the globe.
- The Regulatory and Ethics Working Group, presented by Bartha Knoppers and Kazuto Kato, is currently focused on developing a code of conduct for genomics and clinical data sharing. This, of course, will happen in close collaboration with all other working groups to ensure that their guidelines and standards are compatible with the code of conduct.
- Standardized APIs facilitate compatibility, while at the same time leaving open the concrete implementation behind the API, giving programmers the chance to find optimum data representations for the task at hand. The Genomic Data Working Group, presented by David Haussler, has so far identified four key areas in which current standard file formats (such as BAM and VCF) should be replaced with APIs going forward, and is working closely together with the custodians of current standard formats (BAM, VCF, CRAM) to ensure that any relevant information which can be captured in these formats is included in the APIs:
- Reference variants
- Read data
- Expression, methylation, and other epigenetic data
- The Security Working Group, presented by Paul Flicek and Dixie Baker, has decided to focus its work on the interface of sharing genomic and clinical data, rather than prescribing security measures “behind the doors” of sharing entities. The current goal is to provide a guidance and principle based framework to facilitate trust between data sharing entities.
- The Clinical Working Group, presented by Kathryn North, provided some examples of currently used ontologies (such as Phenotips) and how it plans the integration going forward. Rather than “invent” a new format, the main goal of the group is to develop a unified ontology into which existing structures can be mapped. This is also expected to facilitate adoption by physicians, who have only recently become used to adopting the current standards.
Perhaps the most impressive aspect of the meetup was that it managed to gather so many high-profile experts from around the world to engage in a lively and concrete discussions around these talks. Though many of the alliance’s partners could be viewed as competitors to some extent, the feeling in the room was that everyone had come together because the fundamental ideas and mission of the alliance excited them.
Certainly, the atmosphere for such an initiative is “in the air,” and it should be noted that the alliance is not the first initiative of its kind. Existing data sharing initiatives, such as the Human Variome Project, have somewhat similar ambitions. Several vignettes from existing data sharing initiatives at the meeting, such as the International Cancer Genome Consortium and ELIXIR , indicated that sharing of data is a recognized problem that has been addressed in many ways already. It was encouraging to see the working groups acknowledge the contributions of these initiatives and express the desire to learn from them and integrate existing best practices.
Six months to produce first results
We were honored to be among the first 18 industry partners to join the Global Alliance. The Alliance has taken up an important, but also somewhat daunting task: Ultimately, success of the alliance will be measured by the adoption of its standards. To achieve this, these standards will obviously have to add value to existing frameworks. But it also means these standards will have to be developed quickly to stay ahead of ongoing developments. Coming up with something useful, quickly, that is at the same time simple and actionable, is no easy feat.
The meeting in London had many of the right people with a lot of energy and expertise gather at the table. The next meeting will be in six months. By then, it will be crucial to have developed some initial, tangible results. We’re certainly excited to be part of this ambitious endeavor and will be happy to contribute all that Seven Bridges has learned over the past five years.