Cloud ‘Commons’ would accelerate biomedical research

Cloud ‘Commons’ would accelerate biomedical research

The National Institutes of Health aims to launch an electronic “Commons,” a community-controlled cloud infrastructure that would support collective uses of computing, storage and data for biomedical research by NIH and its academic and industry collaborators.

NIH’s Office of Data Science is in the process of prototyping the Commons, which it described as “a shared computing environment that takes advantage of emerging cloud computing and existing high-performance computing resources.”

In a sources-sought notice for contributors, NIH noted that its biomedical and health sciences research was fast outgrowing the means – through journal articles, scientific conferences and simple networks – to circulate data to the research community.

Instead, biomedical researchers need a new mechanism that is scalable, straightforward to use and could take advantage of emergent cloud and high-performance computing resources, said NIH.

“Transforming this data into knowledge requires that it be made broadly accessible, easily computable and readily sharable among those that are qualified to evaluate it,” according to the notice.

 “The Commons is a pilot experiment in the efficient storage, manipulation, analysis and sharing of research output, from all parts of the research lifecycle,” wrote Philip E. Bourne, associate director for data science at NIH, in a blog post.

The Commons would act as a repository of public data used regularly for biomedical computation as well as provide access to informatics tools.

Over the long term, NIH said, the Commons might be implemented as a federation of public and/or private clouds and high-performance computing centers.  As NIH described it, the Commons would constitute an open set of services that would allow researchers to coexist and access new, existing and ongoing projects.

To accommodate the openness and range of the shared services project, NIH is proposing a non-standard business model for it, starting with a cloud federation that meets standards defined by NIH for connectivity, ease of use and other factors.

The so-called “conformant clouds” could be public providers, companies that use public providers or high-performance computing centers at national labs or research universities.

“Once data or software are in one of these conformant clouds, data sharing becomes much simpler; investigators merely need to authorize access to the research object to the proposed collaborator, who could then bring computational resources to bear (in the relevant cloud)as needed,” NIH explained.

Rather than directly fund content providers, the model calls for NIH to distribute Commons credits to investigators via a third-party reseller under contract to the NIH.  The credits would then be used by researchers for storage and computing by the cloud provider of their choice. The provider would bill the reseller for services used, up to the value of the credits.

If investigators needed more resources for further research they could use other sources of funding, including grants and philanthropy to pay for services.

“This model matches supply to demand and provides resources where needed (with individual investigators) and provides for a robust market where vendors compete to provide the best services to the research community at the lowest possible price,” according to NIH.

NIH said it is looking to identify a reseller of cloud services to manage the distribution of credits and payments as well as organizations that can provide conformant cloud services.

Responses are due Jan. 19, 2015.

About the Author

Connect with the GCN staff on Twitter @GCNtech.

inside gcn

  • machine learning

    Mitigating the risks of military AI

Reader Comments

Please post your comments here. Comments are moderated, so they may not appear immediately after submitting. We will not post comments that we consider abusive or off-topic.

Please type the letters/numbers you see above

More from 1105 Public Sector Media Group