YouTube Facebook LinkedIn Google+ Twitter Xingrss  

The Uncommon Information Commons


By Salvatore Salamone

July | August 2006 | Like most life scientists, researchers at the University of Pittsburgh’s Graduate School of Public Health spent a great deal of time managing data.

As is the case in many labs, data were stored in Excel spreadsheets that were e-mailed to colleagues. As such, much time was spent formatting and preparing data for analysis. And when data was shared, there were also difficulties tracking any changes to the data to ensure everyone was working with the same information.

“I was working on many projects with lots of colleagues; it was hard to synchronize all of the information,” says Michael Barmada, associate professor of human genetics in the Graduate School of Public Health. “I started looking for a data management solution.”

Barmada notes, “We were dealing with huge amounts of data, and we were looking for ways to fit all types of data together. We were getting swamped.”

In his quest for a solution, Barmada heard a talk by Josh Knauer, director of advanced development at MAYA Design, about a new type of database called the Information Commons, a peer-to-peer system that allows many people to securely post and share large amounts of disparate data. In today’s vernacular, it’s something like a wiki database where many users can contribute data, designated people can edit or change the data, and in this case, the originator of the data can selectively control who sees what data.

Similar, But Different
The Information Commons came out of work done by MAYA Design over the past 15 years that was funded by $50 million in research contracts from a number of government agencies including DARPA (Defense Advanced Research Project Agency), the Department of Defense group that played a large role in the early development of the Internet.

The funding was to look into the flow of information in distributed computing environments, something that is called information liquidity.

In a 2005 white paper titled “Designing the Future of Information: The Internet and Beyond,” the consultancy Harbor Research claims the Information Commons is a disruptive technology, meaning it is “not incremental improvements, patches, Band-Aids, or new flavors of what we already do.”

In Barmada’s case, his work with MAYA Design has led to the development of what is being called the Genetic Information Commons (GIC).

Many of the elements of the Information Commons approach will seem familiar to life scientists. For instance, at the heart of the MAYA approach is something called a u-form, a collection of data attributes and their values. Each u-form has a universally unique identifier (UUID).

On the surface, this sounds similar to Semantic Web technology, which uses a uniform resource identifier (URI), metadata about data elements, and information about relationships between data elements. But Knauer says there are distinctions. For instance, UUIDs are not location/server specific (like many Semantic Web URIs). This helps to ensure real data liquidity and means you can get to data even if some of the servers are down.

Additionally, Knauer notes the MAYA approach is more a system for storing and massively distributing data through a peer-to-peer network of repositories. In this system, u-forms get “shepherded throughout the peer network,” says Knauer.

The real power of the Information Commons approach is the ability to securely share data. All data are encrypted and digitally signed when posted. Thanks to the encryption, a researcher can post all of his or her data, but then can selectively designate who can view it. Additionally, the digital signing of the data serves as a check to ensure that none of the data has been altered.

Because all data are represented in u-forms, it is easier to work. For instance, while a database might include an entire genome, a researcher can easily select the data associated with a single gene for analysis.

This helps change the way Barmada and his group do their work. Using the GIC, “We’re not asking how do we integrate this data, but how do we analyze this data?” he says.

For those wanting to try the Information Commons approach, MAYA Design will offer the basic software for free. If specific tools are desired, MAYA typically partners with an organization where both apply for a research grant or contract to fund the development.

Click here to login and leave a comment.  

0 Comments

Add Comment

Text Only 2000 character limit

Page 1 of 1


For reprints and/or copyright permission, please contact  Terry Manning, 781.972.1349 , tmanning@healthtech.com.