$1.7 million to open science’s Friend

Two honors this summer have burnished the reputation of Seattle scientist Stephen Friend for his pursuit of what is called open science.  On top of White House honors in June, he received a grant last week that adds to his credibility as someone trying to change the culture. The phrase “open science” has dozens of meanings, but in Friend’s case it is primarily about opening up data that medical researchers once kept hidden from each other. He calls it a “geek sandbox.”

Just a few days ago, he won a $1.7 million grant from the Robert Wood Johnson Foundation.  The money goes to a project called Bridge, more about that later.

Friend has spent decades being an academic (at Fred Hutchinson Cancer Research Center) and working at the giant pharmaceutical company, Merck, but now he’s director of his own pioneering effort called Sage Bionetworks.

When he explains this passion for open science, he talks from his experience. He talks about trying to make data work for patients and not just to help scientists with tenure and startup companies.  He titled one of his speeches: “Dreaming of tenure and IPOs while patients die,” at a conference in 2012.

He speaks critically of the “medical industrial complex” getting in the way of helping patients. He’s building tools that are designed to let researchers share and network with patients to try to help solve some puzzles about chronic diseases. Already, they’ve had some success with modeling which women with breast cancer might respond best to certain medicines.

Friend was honored at the White House Office of Science and Technology in June as one of the leaders in open science. He announced new projects on Alzheimer’s disease and rheumatoid arthritis that may help predict which patients will respond to certain medications. Sage Bio’s work on modeling risk for breast-cancer patients was called a “geek sandbox” by one writer, who said the results (a new model) are less important than the building of a system where researchers collaborate and compete at the same time.

Friend’s particular passion is also a place where two Seattle strengths overlap, like Venn diagrams. Seattle is strong in software, everybody takes that for granted. Besides the behemoth of Microsoft, Seattle has hundreds of other companies exploring how to use algorithms to uncover signal patterns in the noise of data.  Among the newest of those companies are Socrata and Tableau Software.  Seattle is also strong in genomics and proteomics, and the study of diseases as complex systems that require millions and millions of data points to begin to understand. Software development is ahead of medicine in seeing the wisdom of open source design, he points out.

Friend believes that allowing medical researchers to see each other’s terabytes of data will allow patterns to emerge that spell answers for some diseases. In fact, he believes the institutions and the foundations that donate money must adhere to five concepts – including sharing with patients – or they will fail at finding answers for chronic diseases.

In a video discussion with the patient-advocacy group, Faster Cures, he listed these five rules for success.

1. Make use of massive genomics, proteomics and other “omics” data

2. Put that in the cloud so thousands can access it from anywhere and doesn’t use your server space.

3. Use network modeling of disease.

4. Give patients the ability to send you data (for example, measuring their health using a cell phone)

5. Use open social media to involve patients and share with other researchers instantly.

Friend’s colleague, Thea Norman, talked to us about coming to Seattle to join him and how she will help leverage the RWJF money to make patients bigger partners in research. Bridge is designed to help groups of patients donate their own medical data in a way that provides a basis for discovery for researchers. For example, melanoma patients have already donated photographs of their own skin abnormalities to help build a database that scientists can scan.

“He’s visionary and inspiring,” she explained, of her decision to leave a job in San Diego and move to Seattle to work with Friend. Her title is director of strategic development.

“We get the cream of the crop of software engineers,” she explained of the special strengths that Seattle brings to the open science work. Some of what Bridge will do is build the web-based platform where patients can volunteer their data, or read and download consent information. (John Wilbanks, also at Sage, is one of the designers of what is called Portable Legal Consent.)

The next blog in this series will explore some of the patient advocacy that is driving open science. As Friend says, patients are the key to the enormous energy needed for change.


Digging – the best way to learn science

I remember meeting Maureen Munn, PhD, for the first time, almost 20 years ago. As neighbors and mothers, we often took our six children to the beach together. Her passionate belief in letting children dig and experiment to find out about the world was obvious. Let’s just say – they got dirty. Really dirty.

Now that she has won a prize from Science magazine for designing a curriculum around digital digging in a database, I am grateful her methods brought national praise. Munn has been remarkable in discovering better and better ways to help students dig, and serving the cause of science literacy for a long time from her position as director of education outreach at the University of Washington Department of Genome Sciences.

As a scientist, she has championed the hands-on explorations that are also labeled “inquiry science.” On July 25, she and her team of collaborators were honored with an award for their curriculum about smoking behavior that lets students “dig” through survey data about smoking behavior.

She explained to the National Science Foundation that her curriculum idea grew out of genetic sequencing she did with high-school students focused on finding variation in one of many genes that might be associated with nicotine addiction. Teachers and students told her they wondered about how that might relate to behavior and addiction.

“They were very interested in learning about those topics because they did affect them,” Munn said, in an interview with the NSF.

Over a period of time, she and colleagues collected some survey data from about 300 adults who smoked regularly and others who tried it and quit. They also collected blood samples from the survey participants, purified their DNA, and typed their genomes at three regions known to be associated with smoking.  The resulting database of data became the rich material in which new students could “dig” for answers. Was one thing caused by another or just correlated? How significant were comparisons?  What numbers could they trust?

Students use the curriculum “Exploring Databases” to learn the fundamentals of statistics and the limitations of asking people survey questions.

“Part of their learning experience is recognizing the shortcomings of the study,” said Munn. “The important thing is that students are able to propose their hypothesis, to be able to say whether the data supports it and to defend their results — or to reconsider them.”

When I called her to offer congratulations, Munn said she hopes teachers across the country use the free materials at the project website. “If we have done a good job, new teachers will be able to self-learn how to use the curriculum at our website,” Munn explained. Of course, the grant that funded “Exploring Databases” expires in less than 60 days. She hopes the publicity from the award may keep teachers visiting the curriculum website, even though no new work will be done on the site itself.

“Whether using a micropipette or a database, students develop a better understanding of science if they experience the practices of scientists,” she said. That includes testing ideas, trashing what doesn’t work and finding the evidence to defend a claim.

Don’t worry too much. Munn is busy on a new project about nematodes and interactions of their genomes with the environment.



Small words – Town Hall talk for Ignite Seattle

Some people in science like to use big words.

You know what I’m talking about. Climatology, prehistoric, proteomics. Or phrases that tie us up in knots.

I’m a fan of small ordinary words. Words that all of us use every day. The great cartoonist Randall Munroe of xkcd created a cartoon about labeling the parts of a Saturn V rocket – using only the most common 1,000 words in English. Hint: upgoer is the best way to describe a rocket if you can’t use the word rocket.

His colleague, Theo Sanderson, thought this idea was so much fun that he created a text editor gizmo where you can try it for yourself – and immediately get the red pen if you accidentally use the wrong word. About 180,000 people have already used the upgoer gizmo and many people post their creations to Twitter using the hashtag #upgoverfive.

Because I am a huge fan of this idea of distilling your science into ordinary words, I volunteered to give a public talk about upgoer at Seattle Town Hall as part of the Ignite Seattle nonprofit project.  The video of that talk is above.

That audience laughed a lot about the examples I gave from a competition. You can read more about that competition at my earlier blog post – “Hair Having animals.”

But here are some thoughtful stories about science writing:

Using metaphors:


Write like a human:
American Association for the Advancement of Science advice
A scientist makes fun of scientist writing