Knowledge-sharing platforms emerge from life science research collaboration

One of the hottest topics at life science conferences these days is collaboration. For budgetary reasons, pharmaceutical companies that 10 or 15 years ago would have handled every aspect of research and development in-house have externalized those services to academic partners and outsourced service providers.

"The big pharmas are now under pressure to squeeze out inefficiencies. One response is mergers; another is becoming like Hollywood studios. They just focus on the marketing while outside groups handle other aspects of research," says Barry Bunin, CEO of Collaborative Drug Discovery, a for-profit company offering a cloud-based knowledge-sharing platform for life science research. The industry is seeing the economics of specialization. He explains, "Whoever is best at one particular aspect of the research chain focuses on that and then hands it off."

Alan Louie, research director for clinical development strategy and technology at IDC Health Insights, says the push for greater collaboration is an ongoing trend in the industry, largely driven by the business needs of the top pharmaceutical companies. He has studied the impact of those changes on how data is shared.

In a blog post, Louie writes, "Effective collaboration will be key in this new, distributed future, bringing new challenges to controlled data access and security, as well as new complications in the creation and protection of intellectual property. Clouds and portals are the new ecosystems and interfaces in this new world, and team approaches have already become the new norm as the complexity of efforts have rapidly exceeded the capabilities of any one group or one organization."

A KM challenge

Generic software and platforms designed to facilitate enterprise collaboration, such as SharePoint, aren't geared to managing complex scientific information effectively. Life science organizations have been hampered by a lack of interoperability between their internal systems to manage collaborative projects. That has led to the development of several efforts to create knowledge management platforms within enterprise boundaries and sometimes across them.

For instance, Eric Perakslis, former senior VP of IT at Johnson & Johnson Pharmaceuticals R&D (and now chief information officer at the U.S. Food and Drug Administration, fda.gov), faced a challenge: Johnson & Johnson researchers stored data in more than 100 database silos that could not easily be brought together. In 2009, Perakslis launched a KM project called tranSMART that involved open source tools and a private cloud to build a common enterprise repository to support translational research.

Later, Johnson & Johnson made the decision to transform tran-SMART from an internal project to one that was more broadly open and available for pre-competitive cooperation. "This type of platform has the potential to allow for pre-competitive data sharing on translational research," says Brian Athey, professor and inaugural chair of the Department of Computational Medicine and Bioinformatics at the University of Michigan Medical School. "It can keep companies from replicating the same work and going down blind alleys, and it can avoid needless duplication of software systems."

Consortia focus on common projects

In Europe, a public-private partnership between pharmaceutical companies and the European Union government called the Innovative Medicine Initiative (IMI, imi.europa.eu) has been formed. Pharmaceutical companies make in-kind contributions of research and informatics efforts; public monies are used to support academic collaboration. "Teams of pharmaceutical companies and academic partners create consortia to work for five years on projects like ‘finding a biomarker,' " explains Michael Braxenthaler, president of the Pistoia Alliance, a nonprofit group of life science companies and academic groups that aims to lower barriers to innovation. "When that biomarker is found, information is shared among the partners or more often made public."

Those collaborative groups in IMI have the same problems that Johnson & Johnson has: how to store, manage and share data, Braxenthaler says. In late 2011, the eTRIKS project was set up to develop a KM infrastructure to support all IMI projects. eTRIKS builds upon the tranSMART platform.

In parallel, Johnson & Johnson released tran-SMART into the open source community. Today, the Pistoia Alliance and University of Michigan are leading efforts to build a sustainable ecosystem around tranSMART as a global open source data-sharing and analysis platform. It enables clinical and translational research that includes a community of pharmaceutical, diagnostic and other for-profit companies, not-for-profits and patient advocacy groups, academics and government organizations supported by value-added service providers. Athey and Braxenthaler are serving as co-CEOs for the foundation.

Creating a vibrant community

The tranSMART platform combines a data warehouse with access to federated sources of open and commercial databases upon which analytical tools are integrated. TranSMART builds upon tools and capabilities funded by the National Institutes of Health (NIH) through the National Centers for Biomedical Computing (NCBC, ncbcs.org) program including Informatics for Integrating Biology and the Bedside (i2b2), the National Center for Biomedical Ontology (NCBO) and the National Center for Integrative Biomedical Informatics (NCIBI).

Researchers might use tranSMART to search across multiple data sources to find associations of concepts, such as a gene's involvement in biological processes or to compare biological processes and pathways among multiple data sets from related diseases or even across multiple therapeutic areas. Besides getting the tranSMART Foundation fully established as a legal entity, Braxenthaler says another goal for 2013 is to continue to work on the core architecture and create a vibrant open source development community around innovative tools to analyze data.