NCI picks Amazon, Microsoft for cloud data project

NCI's Shady Grove campus

The National Cancer Institute (NCI) is teaming up with Amazon ($AMZN) and Microsoft ($MSFT) to work on genomic data platforms. The idea is to tap into the tech giants’ cloud computing knowledge and capabilities to create a sustainable model for sharing cancer genomic data with researchers.

NCI has spent several years advancing its cloud computing ambitions, both to ensure data is widely available and stop computing constraints from slowing research. In 2014, the institute signed up The Broad Institute, The Institute for Systems Biology and Seven Bridges Genomics for its Cancer Genomics Cloud (CGC) Pilots. And, with the Cancer Moonshot program advancing, NCI has an increasingly clear mandate to ensure researchers can capitalize on the genomic data boom.

The pact with Microsoft and Amazon Web Services--the cloud computing wing of the e-commerce powerhouse--is intended to complement existing NCI initiatives. Specifically, NCI is hoping its new partners can help build a “sustainable model” for making cancer data available to researchers through the CGC Pilots and Genomic Data Commons. Shifting the computing burden from individual researchers to central cloud platforms should widen the reach of the genomic datasets.

“By providing avenues for cloud storage, genomic analysis, visualization and computation, the collaborations help maximize the ability of researchers to mine cancer data for answers to reduce the impact of cancer on the American people and the world,” the White House wrote in its summation of the pact.

The White House unveiled NCI’s tie up with Amazon and Microsoft as part of a report on the progress of the Cancer Moonshot program. Multiple aspects of the update touched on biotech IT. The Joint Pathology Laboratory at the Department of Defense is working to digitize and share its repository of 34 million pathology samples. By pairing the resource with artificial intelligence, the group thinks it has a shot at improving cancer diagnosis.