Michael J. Fox Foundation Brings Parkinson's Data to tranSMART

By Bio-IT World Staff 

May 15, 2015 | Ken Kubota, Director of Data Science for the Michael J. Fox Foundation, is a big fan of the Framingham Heart Study. Since the 1940s, this project has tracked the incidence of cardiovascular disease among three generations of residents of Framingham, Mass., collecting an incredible amount of data about the lifestyle factors that put people at higher risk for heart failure.

“We know everything about heart disease from that study,” says Kubota. Many of its headline findings, like the fact that a healthy diet and regular exercise can protect against heart disease, now seem so obvious it’s almost hard to believe they were ever unknown. The Framingham study has also turned up genetic and pharmaceutical factors that play a role in heart conditions, like a protective effect of taking daily aspirin.

At the Fox Foundation, Kubota says, “We want to do the same thing for Parkinson’s disease, but we don’t want to take 75 years.”

Since its inception in 2000, the Fox Foundation has become a leading player in research on Parkinson’s disease, providing hundreds of millions of dollars for both research grants and its own large-scale studies. Over the years, these studies have amassed huge amounts of data, which is supposed to be freely shared and integrated to create a more complete picture of the biological factors underlying Parkinson’s disease. However, like many non-profits trying to drive basic science, the Fox Foundation has found that making data truly open is easier said than done.

This April, at the Bio-IT World Conference & Expo in Boston, the Michael J. Fox Foundation received a special Judges’ Prize in Bio-IT World’s Best Practices Awards for a major effort to not only share its data with a wide variety of collaborators, but also make sure that data can be readily understood and analyzed across institutions. To accomplish this, the Fox Foundation adopted tranSMART, an open source platform where data, curated in highly interoperable formats, can be accessed through a web portal and merged with outside groups’ proprietary datasets. The tranSMART platform, maintained by the non-profit tranSMART Foundation, is a favorite venue for big pharmaceutical companies to analyze their data, furthering the Fox Foundation’s aim to make it easier and more financially viable to develop new therapeutics for Parkinson’s disease.

The Fox Foundation’s ongoing project to make data available through tranSMART, carried out in collaboration with Thomson Reuters, also involves the creation of innovative new research tools that are compatible with the platform. Currently, the Fox Foundation is placing in tranSMART all the data from its landmark Parkinson’s Progression Markers Initiative (PPMI), a study of hundreds of individuals that includes a rich collection of biological variables, including blood and cerebrospinal fluid samples, MRI and PET scans, and surveys of physiological symptoms, all of which may help discover early warning signs, drug targets, or new clinical subtypes in Parkinson’s disease.

“We can look across studies and begin to piece together a natural history of Parkinson’s disease,” says Kubota. “What correlates? What moves together with the different symptomology of the disease, and the different biologic markers we have information on?”

Rich Data 

TranSMART began its life as a platform for research inside big pharmaceutical companies, first developed within Johnson & Johnson and later expanded with support from Pfizer, Sanofi, and Takeda. By 2013, however, the effort had grown so large that it was spun off as an independent foundation. That foundation had a mandate to make tranSMART’s code a completely open stack, bring on board industry standard data analysis tools, and reach a broad user base that would enable the sharing of data across many organizations.

This transition brought with it new kinds of users flocking to the platform. Pharma is still a major adopter, but the majority of tranSMART users, like the Michael J. Fox Foundation, now come from the basic research side. (One of these users, the U-BIOPRED Consortium focused on asthma, received its own Best Practices Award from Bio-IT World last year.)

The tranSMART Foundation has had to adapt its priorities to the needs of this very different community. “One of the key things that we’ve seen is that pharma companies are typically consumers of content, and non-profit foundations and academic medical centers are typically producers of content,” says Keith Elliston, CEO of the tranSMART Foundation.

That’s certainly true of the Michael J. Fox Foundation, which wants to place large amounts of anonymized patient data from the PPMI in a web portal that researchers can easily access. “The point of a clinical trial is to get at the data, and the only way we’re going to be able to solve this disease is when the minds of the world have access to it,” says Kubota.

At the tranSMART Foundation, a lot of recent work has gone into dealing with new data types, as next-generation technology expands the universe of variables that can be measured in any given study. “We see this convergence of big genomics data with big sensor data, and the third thing we see is that the clinical trial is moving from the hospital to the home,” says Elliston.

The Fox Foundation is well ahead of the curve on these trends. It has been an early adopter of large-scale genetic sequencing, even partnering with the consumer genetic testing company 23andMe to offer free testing kits to more than 10,000 Parkinson’s patients. The Fox Foundation has also been enthusiastic about wearable sensors, which promise to provide a more objective, quantifiable look at the severity of Parkinson’s symptoms than traditional surveys. The Foundation is planning to distribute wearable devices or movement-tracking apps to members of the PPMI and other studies, to help analyze how symptoms change over time.

“Hopefully, along with this very rich biological data, we’ll have continuous, objective measures of their state of Parkinson’s disease, for a very long time,” says Kubota.

Now that the Fox Foundation has adopted tranSMART, future projects tying all this data together will have a better-defined set of standards, and new data collection and analysis tools can be built with integration with tranSMART in mind. For example, the Fox Foundation is now building a patient registry called Fox Insight, where individual patients can choose to deposit as much of their personal information as they like, from their medical histories and medication regimens, to the results of genetic tests or continuous data streams from movement tracking apps. That entire project is being designed to be compatible with tranSMART, providing a ready venue for researchers to access data that patients share with Fox Insight.

This kind of forward-thinking resource for Parkinson’s research, which can readily integrate different data types and provide built-in consent protocols for patients, was part of the reason the judges of the Best Practices Awards considered the Fox Foundation’s tranSMART implementation a model for other groups focused on specific disease areas.

Common Ground 

While the Fox Foundation wants to make all its data as public as possible, not every organization working on Parkinson’s disease can be so open.

“There’s always going to be proprietary data,” says Rudy Potenzone, a long-time advocate for interoperable data in the life sciences who currently serves as the tranSMART Foundation’s Vice President of Marketing. “There’s going to be data you can’t publish for one reason or another, or you don’t want to because you’re using it to design a drug, and in those cases you want to mingle it with all the open data you can get.”

Part of the attraction of tranSMART for the Michael J. Fox Foundation is that the platform is already a standard tool for pharmaceutical companies. Placing public data there makes it easy for Parkinson’s programs within pharma to not only access the results from projects like the PPMI, but also combine it with their own proprietary data to get richer insights.

This was a major consideration while the Fox Foundation was searching for its common data analysis platform, according to Kubota. “This is the platform that pharmaceutical companies prefer, and what they use in their R&D efforts,” he says. “It’s just a natural, obvious place for the data.”

That flexibility to integrate datasets in tranSMART also provides an opportunity to consider new connections between previously disparate fields of study. As data from the PPMI is placed in tranSMART, the Fox Foundation also plans to standardize that data to be directly comparable with studies of other neurodegenerative diseases, starting with the Alzheimer’s Disease Neuroimaging Initiative.

Kubota suspects that both commonalities and differences between these diseases could provide a more precise picture of the biological pathways that cause their distinct symptoms. “When you look across Alzheimer’s disease, Huntington’s disease, Parkinson’s disease, there’s protein aggregation involved,” he says. “They aggregate in different areas of the brain, and they’re different kinds of proteins, but are these diseases really all that different from that perspective?”

Drawing connections across disease categories has traditionally been difficult, because even studies that collect the same kinds of data on these different patient groups have not used the same standards or formats to store and analyze it. Even retrieving data from a partner who is willing to share, and curating it to a point where it can be understood and compared, is an inefficient process that can take days or weeks. In a common environment like tranSMART, this barrier is much easier to overcome.

“We can’t afford not to make use of all the data we can possibly make use of,” says Potenzone. “And that’s what the Michael J. Fox Foundation is trying to do… The next set of people working on the project should not suffer because they don’t have access to all the information.”

The tranSMART Foundation played a role in several award-winning projects showcased at this year’s Bio-IT World Conference & Expo. The tranSMART platform version 1.2 received a Best of Show Award in the category of Informatics for outstanding new life sciences products, and the winner of the conference’s poster contest, Ward Weistra of the Hyve, used tranSMART to analyze and visualize cell line data.