NEW YORK (GenomeWeb) – A year after its official launch, the Open Humans project, which allows individuals to donate their own data — including genomic information, microbiome profiles, and activity tracking data — for research, is ready to grow its membership and engage more research teams who want to access its data.
"We really want to lower the barrier for research studies to work with people who want to donate their data," said Madeleine Ball, co-founder of Open Humans and director of research at PersonalGenomes.org. "We're not a study, we connect people to studies."
Open Humans kicked off a year ago with $1 million in grant funding, provided in equal parts by the John S. and James L. Knight Foundation and the Robert Wood Johnson Foundation. The project was created by researchers at Harvard University, New York University, and the University of California, San Diego, and is run by PersonalGenomes.org, which also administers Harvard's Personal Genomes Project.
Through an online system, individuals can upload health and other data that they obtained by participating in research studies, such as the PGP, GoViral, or American Gut; by using commercial services, for example 23andMe, AncestryDNA, or Illumina's Understand Your Genome; or from fitness tracking devices and apps, such as Fitbit, Jawbone, or Runkeeper. In addition, they can upload any data about themselves, for example medical tests, as a so-called "data selfie."
The project has no formal arrangements with any of the data providers because it is the users who submit the data. "We did not enter into a collaboration with them," Ball said, "but we're unlocking their data to be used for research purposes."
Open Humans hosts the data and processes it where appropriate, for example stripping a user's name off a genotyping data file from 23andMe. Whenever available, the project uses application program interfaces (APIs) to access the data, which are provided by many of the fitness trackers and by some research studies. That way, participants only need to authorize the use of the data once and it will be updated automatically.
Last year, Open Humans also launched its own study, called "Public Data Sharing," where project members sign up to share data of their choosing publicly to make them even more accessible to research. Many of its members have joined this effort. "Some people are OK with sharing their genome but they don't want anyone to know where they live. Other people are fine with sharing GPS maps of every run they've ever done but they don't want anyone to know their genetic data," Ball said. "So it's really up to the user what they want to make public."
In total, more than 2,000 people have signed up for Open Humans so far, including many from the PGP, which has recruited almost 5,000 participants over the past 10 years or so. Going forward, Open Humans leaders are hoping for a tenfold growth in membership. "We have the infrastructure in place to support all of that," Ball said.
The project also plans to support more health-related data, for example electronic health records, but those will be more difficult to handle, Ball said. "We've tackled the easier data first."
Researchers wanting to use data from Open Humans members can create a study on the site and recruit participants. The project recently streamlined the process for doing this: After entering information about their proposed project, researchers can connect to up to 10 test users — usually people they know —, download data authorized by these users, and send them messages. After this test run, they can apply to open their project to all Open Humans members.
In addition, researchers can use the project's Public Data API to work with data members have shared publicly.
Research studies need to be approved by an institutional review board, Ball said, though Open Humans also plans to allow projects that are not research studies, for example data visualization that would enable members to interpret their data in new ways. "This is all going to be case by case — we don't guarantee approval for anything but we are very transparent about projects coming for approval," Ball said.
Open Humans also has access to legal and IRB experts who can advise researchers on their studies and how to return data to participants without requiring approval from the US Food and Drug Administration. "There are ways for studies to give data access that do not turn them into clinical devices," she said.
Originally, Open Humans required research studies to return results to its members, but "we're shifting away from that," Ball said. While the project strongly believes that studies should give participants access to raw data, and will continue to promote studies that do so, "we recognized that not every study is going to produce data that's useful for reuse, or even to the individual," she explained. For example, some studies just consist of a survey, and getting one's own survey responses would not be all that interesting.
At the moment, Open Humans has fewer than 10 research teams lined up who are interested in using its members' data, but the project is hoping to attract more studies in the coming months and years.
Over the past year, the project has largely focused on discovering "what works and what doesn't work," Ball said. In general, it has found that people are happy to share, as long as it is easy to do so. "People want to share — the biggest barrier for most people is just not knowing which buttons to click," she said. "We always want to be very careful that people don't share inadvertently, but we also need to make it easier for them to share what they want to."
Starting this spring, Open Humans plans to promote itself more widely, for example at the upcoming Genomes, Environments, Traits (GET) conference in Boston next month, which is organized by PersonalGenomes.org.