Stephen Friend, Leaving High-Powered Merck Gig, Lights Fire for Open Source Biology Movement
Stephen Friend had it all at Merck. The lucrative salary, the national media fame, the respect of peers, the power to snap his fingers and command a corporate army of some of the brightest minds in biomedicine. This year, he gave it all up for a startup dream.
Friend, 54, is best known in Seattle as the founder of Rosetta Inpharmatics, the cutting-edge genome analysis company that was acquired by Merck in 2001 for more than $600 million. For the past eight years, he worked on the East Coast as a senior vice president for the pharmaceutical giant, and spent much of his time running the company’s worldwide cancer research. Now he’s back in Seattle, dreaming big again as founder and CEO of a global nonprofit collaborative called Sage Bionetworks.
Sage, as we first described back in March, is attempting to do for biology what Facebook and Twitter have done for social networking, and Linux has for open-source software. Sage is needed because biologists are beginning to see how vast networks of genes get perturbed in complex diseases like cancer, diabetes, and multiple sclerosis. All of this genomic data is too complex for any individual, or team of scientists—even at a place as big and rich as Merck—to fully grasp. Yet researchers scattered around the world are capturing huge volumes of genomic data on their computers, hoping it will someday be fodder for groundbreaking discoveries. Sage wants to get all these biologists to now start sharing their data in a free, open database, so they can pool their brainpower to come up with insights that one person, or team, never would discover in isolation. The ultimate result would be more effective drugs, just like open-source computing can create better software.
Friend has been pushing for this type of collaboration for years at Merck, but it wasn’t really possible inside the proprietary walls of a company. That’s why he had to start something new. He told me how thrilling, and terrifying, it was to suddenly shed all the support he had at a big company, and be back in startup mode, pouring every ounce of his energy into the one thing he’s most passionate about.
“It feels like I’m moulting,” he says.
Since the original story broke in March, many essential organizational tasks have been done. The first checks have been deposited from donors (Friend won’t say how much) out of the original $5 million in commitments. Accounting systems have been set up. The intellectual property that Merck agreed to hand over, that represents eight years and $150 million of work, has been signed over and transferred. Through help from his friend, Nobel Laureate Lee Hartwell, the new nonprofit has set up office space at the Fred Hutchinson Cancer Research Center. Fifteen of the bright minds from Merck’s Rosetta operation have come to work at Sage full-time. Phones and e-mail are up and running.
Sage still isn’t fully disclosing who its supporters are, although the Bill & Melinda Gates Foundation and the U.K-based Wellcome Trust are “stakeholders” that have said they would like to participate as the ground rules are being framed, Friend says. One foundation, the Cure Huntington’s Disease Initiative, has agreed to disclose it contributed some of the seed money to Sage, Friend says.
When I asked him how far it’s come since we first discussed the Sage vision in March, he said, “It became real.”
Like with any startup, there have been bumps. Friend’s co-founder of the new nonprofit, Eric Schadt, left Sage in May to take a job as chief scientific officer of Menlo Park, CA-based Pacific Biosciences. Schadt said this makes sense because PacBio is developing more advanced genetic analysis tools that will be needed to really enrich the Sage database in the future, and that he can benefit both the company and the Sage database of the future by building up both organizations at the same time. So far, Friend says his partner is making good on his promise to spend one day a week working on Sage, and is traveling to Seattle frequently to stay in touch.
But really, Friend wanted to spend most of our interview talking about the deeper vision of Sage, rather than these nuts-and-bolts operational details.
The vision of Sage, Friend says, is to create the first-ever accurate disease models. This is sort of like how engineers can draw up models of how a plane with certain specifications should fly, which tells them a lot of what they need to know before actual lift-off. Having that kind of underlying model to understand basic laws of physics helps scientists interpret data streams they observe later, like when a plane is flying.
The pharmaceutical business, because it is dealing with the immense complications of what’s going on in the human body in real-time, has never come close to any models with this kind of predictive value. Scientists do experiments on animals, but they’re not the same species. Medicine currently relies on lots of what Friend calls “linear measurements,” like LDL cholesterol as a marker of heart attack risk, or Her2 gene mutations to grade the type of breast cancer a woman may have.
Diseases are actually far more complicated than a matter of a single gene or protein being out of alignment, but so far, these are some of the best facts medicine has to work with, Friend says. But now that sophisticated new genetic instruments are getting better, faster, and cheaper, it opens the door to much deeper understanding of the genetic symphony of what’s going wrong in an individual patient, and what kind of treatment might give them a greater likelihood of response. These tools can do everything from sequence the entire 3 billion letters of DNA in an individual in a row, to finding slight variations in the code called SNPs, to seeing the extent to which certain genes are turned on or off in a given sample—what’s called gene expression.
The experiments at Rosetta that gave birth to Sage looked at all these measurements, many of which were published in top scientific journals. Some of the most important work showed how abnormalities in individual DNA caused faulty transcription of RNA, which caused malfunctioning proteins, and connected all that underlying biology to a clinical diagnosis of disease. This sort of holistic data package is hard to find anywhere, and it’s the way Sage wants genomic researchers to think in the future, Friend says.
“We’re going to scour the globe and find collaborators with coherent global genomic data sets with clinical outcome measures,” Friend says.
Friend doesn’t really know how many of these holistic data sets exist in the world, but he’s looking. Whatever it is now, he wants it to grow a lot from today, so what he’s really talking about is a more cross-disciplinary cultural transformation in biology and medicine.
Like any new international organization, Sage needs a broad base of support, and Friend is planning a charter summit for this fall, probably in November. This is going to be a pivotal meeting for the future of Sage, not only to see if Friend can drum up enough support for his vision, but to get people to collaborate on some of those essential organizational details.
These are not small things. Part of the agenda will be to work on ground rules for how intellectual property and data privacy are handled, what incentives researchers have to bare their souls and put their data in the pool, and what standards should be followed to ensure the data is consistent and not full of inconsistencies and errors. Then the user interface will have to be easy to use, or many biologists probably won’t use it.
These are hard questions that Sage will have to wrestle with for months, and probably years, if it’s to fulfill its vision of turning biology into a truly collaborative field of study. But these are the heady days for Friend.
Friend is moving with such energy that he hasn’t even bothered yet to hang a picture on the wall. The day we spoke, he was looking forward to holding a retreat for the Sage team at his home on Stuart Island in the San Juans, and he said some members might just camp out on his yard to get a view of the stars.
An individual on Earth may feel small when looking out at the universe like that, but Friend is drawing some inspiration from small teams of committed people that have shown they can tap into those greater forces. Twitter, for one, has found a way to harness social networks in such a powerful way that it may have altered the course of world events after the Iranian election, Friend says. With a staff of just 43 people.
He’s thinking about how in his small organization, he can light the kind of fire in the world community of biology that could have such an impact. Sage’s success will depend, like Twitter, on getting other people to do the work. And to entice other biologists to participate, Friend is appealing to their desire to discover.
“With new tools comes new understanding, and paradigms fall,” Friend says.