GitHub: an open source tool which allows developers to write their software privately, then upload it and invite others to collaborate, all while being able to track each change made to the source code. In a similar way, precisionFDA could enable researchers from one lab to do their work privately, upload sequencing information and share it with whoever they choose, and invite researchers from another lab to try to reproduce those results and compare. Mistakes could help folks learn from one another, and produce more accurate tests.
“By contributing the right platform there’s a real opportunity to have the key stakeholders reach as much consensus as possible, and provide a trusted forum,” says DNAnexus chief medical officer David Shaywitz.
Will testmakers, researchers, and others embrace the effort? Kass-Hout acknowledges this is one of the biggest challenges—the project will go nowhere if an active community doesn’t form and populate precisionFDA—but says he’s encouraged that “pretty much everybody” he’s talked to support the concept. Another challenge is keeping on a “very, very tight timeline” of launching a beta version of precisionFDA by sometime next year.
“This is a new terrain,” Kass-Hout says. “The only way it can all work is if we have a really large collaborative effort.”
For DNAnexus, meanwhile, this isn’t an exclusive, iron-clad deal. It’s an R&D pact that gives the company the first shot to build precisionFDA—and to profit from it if things work out, potentially by charging fees based on usage, according to Serang. If they don’t, or something better comes along, the FDA could always go with another company, Kass-Hout says.
“We have the specifications, we know what it is we’re building now,” Serang says. “it’s going to be our job to produce a very high value site that the community loves to use.”
