Tony Kerlavage on Data Lakes, Data Commons, and Empowering the Research of the Future
At the National Cancer Institute, Tony Kerlavage knows quite a bit about managing very large pools of data. When NCI launched the Genomic Data Commons, it aimed to democratize access to the genomic data in The Cancer Genome Atlas and other sources. Since then, though, Kerlavage points out that our data types and volumes have only grown. Now NCI is taking a “Commons of Commons” approach to link pools of well-structured data. “The more data we can bring together in a well-structured way, the more value it has in the long run,” he believes. He advocates for sharable Python notebooks and reusable R programming, believing significant investments in data hygiene and interoperability delivers more value than simply mining data lakes with artificial intelligence tools—for now, at least. The challenge for researchers, Kerlavage says, is to view their work with an eye to the future: How might someone else use this data going forward?