Sage Bionetworks Moves from Thinking Stage to Doing Stage
Three years ago, Stephen Friend started a nonprofit with a bold vision to speed up drug discovery, which might represent the biggest cat-herding exercise in the history of biology. If it sounds equal parts promising and daunting, it should.
The idea of creating an open-source movement for biology—a field in which scientists and corporations guard their data under lock and key before presenting carefully selected findings—hasn’t caught on like Wikipedia or Facebook. But this week, Seattle-based Sage Bionetworks is organizing its third annual Commons Congress in San Francisco, to keep building momentum for a movement in which researchers share more of their experimental data and models in the open. The conference, which Friend started three years ago with genomics whiz Eric Schadt, will bring together 250 forward thinkers in person, and more online, from academic science, Big Pharma, biotech, government agencies, and patient advocacy groups.
From the beginning, Friend, a former senior vice president at Merck, said he was on a five-year journey to see if Sage Bionetworks could help create a common place online where scientists could pool their data and brainpower to speed up the pace of discovery and drug development. His reasoning was that biology had become far too complex for any lone researcher or Big Pharma company to keep tackling problems in isolation. By coaxing all kinds of players to participate in an open commons, where scientists pool data on everything from DNA to RNA to proteins to clinical observations of disease, Sage is hoping to help researchers connect dots that might not otherwise ever be connected.
There are a million reasons why such an effort might never fly, and Friend has surely heard them all. There’s the technical part (new software needs to be created). There’s legal (consent forms need to be drafted). There’s a cultural aspect (scientific attitudes and career incentives need to change). And don’t forget the financial part (who will pay for all of this?) and the political/ethical dimension (isn’t all this genomic data going to violate someone’s privacy, or open the door for genetic discrimination?).
Entering Year Three, Friend says it’s time for the open-source movement to build on its early work thinking through the opportunities and challenges, and to make headway on getting more collaborative projects up and running.
“The first year was about ‘imagine a world where …,” Friend says. “We were trying to get people to frame the idea of what they wanted. The second year, we were beginning to do some pilot projects. It was about building better maps of disease together.”
Now, he wants to see more pilot projects in various disease settings, with more people participating and sharing their lessons learned. Even though the invite-only event is maxed out with 250 attendees, Friend says he’s confident the Congress will have a much bigger pool of participants this year on the Web. “I think there is a possibility this year of having thousands of people watching the webcast live,” Friend says. “In the past, we had hundreds of people. We have medical students from Pakistan and China waking up in the middle of the night to call into this thing. We have basic scientists in Sweden and the Netherlands getting up in wee hours.”
The grassroots support from graduate students and postdocs around the world is a critical new ingredient for Sage, which already has quite a few prominent allies helping to spread the word. This year’s group of speakers at the Congress includes leading genomics researchers like Harvard University’s George Church and UC Santa Cruz’s David Haussler; Stanford University legal scholar Lawrence Lessig; PatientsLikeMe founder Jamie Heywood; and Rick Klausner, the former director of the National Cancer Institute. Sage also has rallied some sponsorship support from big institutions in pharmaceuticals, software, data, and entrepreneurship.
But as someone who attended the Congress in its first two years, I recall Friend emphasizing how critical it is that attendees commit to taking certain actions in the other 363 days a year they aren’t attending the conference, to keep the momentum going. My pledge was to keep paying attention, and keep asking hard questions about the progress Sage is making on its mission.
So when I spoke to Friend last week by phone, I asked him to list off two or three accomplishments Sage has made since the last Congress, and two or three items of unfinished business.
The first big accomplishment, he says, was in building a common computing platform known as Synapse. There has also been some progress to report on creating disease-specific consortiums for data sharing, including one announced this week to pool data for research into psychiatric disorders. Another key step came when Sage’s Jonathan Derry published “a coherent set of components and principles on working in an open space,” in the journal Nature Genetics.
At last year’s Congress, Sage didn’t really have a computing platform set up that researchers could efficiently dump data into and view it. There has been significant progress since then, with the Synapse system, according to Sage director John Wilbanks. “Synapse emerging from Sage (and thus from Merck/Rosetta) was not unlike Mozilla emerging from Netscape – a technology that had to be fundamentally turned around to be a public resource,” Wilbanks says. “And there’s remarkable work on doing collaborative computational science on the network that’s inside there, killer stuff around provenance tracking, group permissions, all the blocking and tackling you need to really work on big biological data with people in a distributed manner.”
Then there are the roadblocks. The first one Friend mentioned has occurred with the Sage Federation, in which the nonprofit agreed to get a handful of top biomedical researchers to agree to pool their experimental genomic data. Those scientists—including Atul Butte at Stanford University, Trey Ideker at UC San Diego, and Andrea Califano at Columbia University—agreed to join the Sage Federation, only to see legal adoption of the agreements at the various institutions take an additional six to 12 months. “Legal adoption has been slower than we thought,” Friend says.
The second speed bump came with what Friend calls the “non-responder” project in cancer, in which he has been trying to coax pharmaceutical companies to contribute detailed data from the control groups of big clinical trials. Friend, as a former pharmaceutical executive, figured this would be an innocuous way to get pharma companies to contribute some valuable data without divulging anything about proprietary compounds, but even so, he’s found it to be a hard sell. “It’s been very hard to identify individuals willing to fund those projects,” Friend says. “We’ve moved to other projects in fields like RA (rheumatoid arthritis), where there’s more interest in industry.”
The third item, which Sage hopes to present as a success story at this year’s Congress, is the development of a standardized patient consent form, in which patients grant researchers the right to analyze all kinds of genomic, proteomic and other data from their tissue samples, while protecting basic patient rights and privacy. Wilbanks, a director of Sage and a fellow at the Kauffman Foundation, has played a key role in this effort, which requires lots of careful, time-consuming legal and ethical review.
“Structurally, what we’re running is a clinical study that allows people to upload data that they already have about themselves into a technology that then syndicates the data into computational research environments like Sage’s Synapse,” Wilbanks says. “So we’ll show all the parameters, running software, and the consent form as submitted, but we won’t throw open the gates quite yet.” The consent system still needs to complete the review with the institutional review board that is the watchdog for patient safety and ethical concerns.
Sage clearly has a hard-core group of supporters—about 80 percent of the people who made videotaped commitments at last year’s event are coming back this year, Friend says. And there are some new groups getting interested that tend to have specific disease-related concerns, like a post-traumatic stress disorder research group called One Mind for Research, he says.
When asked if Sage is at the point where he thought it would be three years into a five-year plan, Friend said, “I do—I do feel it’s on track.” But a lot of what comes next will depend partly on what Sage does to clear the way, and what people in the biomedical research community choose to do with its platform. “Our mission is to develop flexible approaches that can be championed by others,” Friend says.
Wilbanks sees major progress in the growing number of champions for the cause. “The conversation around open biology is totally different than it was three years ago,” he says. “We started with hope, now we’re shipping code (both software and law) and people no longer think it’s an impossible dream. That’s not to say it’s going to be easy, but that sense of possibility is itself a real achievement.”