Genomic Search Engine One Codex Wins CDC Prize to Catch Bad E. Coli

Xconomy San Francisco — 

San Francisco genomic data firm Reference Genomics says it has the tools to spot nasty microbes fast. The U.S. Centers for Disease Control and Prevention agrees.

With its One Codex search platform, the startup has won a CDC-sponsored “no petri dish” contest to identify a toxic strain of the bacteria E. coli without the time-consuming process of isolating the bug from a patient’s feces or other sample full of diverse microbial life. The $200,000 prize is to help continue development of the platform.

The CDC sponsored the contest not to encourage rapid identification of microbes—there are plenty of tests to do that, said Duncan MacCannell, a senior advisor for bioinformatics at the CDC. Rather, the nation’s top disease detectives are worried because all the new diagnostic products aren’t funneling information about individual cases—“isolates,” as MacCannell put it—back to public health officials.

“The isolates aren’t flowing back into the public health system the way they used to,” said MacCannell. “We rely on those isolates across the country to see patterns and clusters of disease.”

The CDC contest follows in the footsteps of the Defense Advanced Research Projects Agency (DARPA) and private contests like XPrize to stimulate innovation outside institutional walls.

In this case, the Atlanta-based CDC wants better tools to track infection outbreaks. The particular strain of E. coli used for this contest is a good test, because it’s very similar to the normal E. coli present in everyone’s gut and difficult to detect from a complex sample. “To distinguish good from bad directly from a sample is a difficult challenge,” said MacCannell, and the CDC hopes the innovation will extend to detection of other organisms.

“We haven’t finished the conversation with the CDC, but our goal is to have them use our search engine for epidemiology,” said Reference Genomics cofounder Nick Greenfield. He underscored that the prize is no guarantee of future business.

The One Codex search allows researchers to plug in the data from a sequenced sample and find matches in publicly available repositories, such as those at the National Institutes of Health.

Among the entries, One Codex stood out because of its user interface. “Where it shines is ease of use,” said MacCannell. “It’s not only extremely fast and cloud based, but a lot of the alternative approaches required a fair bit of expertise.”

MacCannell said it’s important to make tools like this available to public health officials and providers without bioinformatics chops or, as he put it, “comfort with a Linux command line,” A reference to computer systems that don’t have a friendly user interface.

In essence, it’s the same problem that companies like Google and Yahoo solved in the 1990s and early 2000s to bring Web use to the masses.

“We want to build a large index and the best reference possible,” said Greenfield. “If you look at genetics papers, you’ll see a number of investigators say, ‘I had a patient and a sample, then I did all this work and found it was [microbe] X.’ People are spending the majority of their time figuring out what X is, and they shouldn’t be. They should be characterizing X more deeply, or figuring out why the person got X or how to keep X from spreading.”

One Codex was part of the first biotech class at the Bay Area accelerator Y Combinator, as Xconomy reported last year.

In addition to the small investment Y Combinator makes in its participant companies, One Codex has raised a seed round, said Greenfield, but he but declined to discuss details.