in the 1990s were using coliform tubes to capture bacterial samples. This method was basically able to collect a lot of different kinds of bacteria, which might or might not be harmful. Officials couldn’t tell where the bacteria came from—whether it was from human, cattle, or bird feces. That obviously can make a big difference in how the government might respond to a bacterial contamination, and tells scientists a lot about its potential severity.
During the early part of the last decade, Goodman pursued this environmental sampling dilemma as “a little hobby,” he says. He testified before the state water quality board, and contributed to reports commissioned by the U.S. Environmental Protection Agency and the National Academy of Sciences. The tools of molecular biology were fast becoming more powerful and cheaper, but he saw that it still took months to ask about a specific type of bacteria in a sample, and cost several hundred thousand dollars, Goodman says.
Not much progress seemed to be happening toward his vision of a fast, accurate, affordable technology for identifying bacteria in food and water. But in early 2007, Goodman got a call from his friend John Hulls, the leader of the State Water Resources Control Board’s “PhyloChip Project,” funded under the Clean Beaches Act.
“He said, ‘Corey, you’ve got to go over to the LBNL (Lawrence Berkeley National Lab),” Goodman recalls. Hulls told him, “You’ll be blown away.”
Goodman, you could say, hit it off with Andersen. “It was obvious within two or three minutes of talking to Gary that it was clear. This was it.”
Back then, the PhyloChip was still in a second-generation form that was only able to detect signatures of 8,000 different known bacteria. But even in that lesser form, “my head was spinning” with the potential use of the tool for detecting bacteria in water and food. Later on, he started thinking about human health applications. And the ideas kept coming, as Andersen mentioned “nonchalantly” that NASA and the Jet Propulsion Laboratory had shown interest, along with oil companies looking for bacteria in oil.
Despite the early promise, and demand from Andersen’s academic peers to use the technology in their experiments, Goodman wanted to see the technology get better. The ideal chip would have to get broader, with capability to detect all 55,000 known bacteria, with room for more as more organisms are discovered. But even before that happened, the word was out. PhyloChip picked up some helpful buzz, winning a 2008 Technology Innovation Award from the Wall Street Journal in the environmental category. By 2009, about 30 different academic labs were using the technology, and publishing a string of papers enabled by the technology, Goodman says.
Goodman figured it was time to start tapping his Rolodex in May 2009, to turn this idea into a company. He called Kreiner to tell him about the opportunity, and Kreiner interrupted him mid-sentence. “Corey, you’re talking about Gary Andersen’s work, aren’t you,” Kreiner interjected.
Then it was a matter of finding the right intellectual property attorney, negotiating for the license from Lawrence Berkeley, recruiting Warrington, crafting the business plan, and recruiting investors. Goodman kept working the recruiting side of things. He proudly noted that he attracted Peter Gleick, an internationally-known water expert based in Oakland, CA, and a recipient of a MacArthur “genius grant” for his work in the field. Matt Winkler, the founder of Ambion, a molecular biology tool company, brought more instrument and startup experience to the board, as well as some of his own money to invest.
The first close of the angel round came in April. The development of the technology was done, and PhyloTech is ready to offer the service to customers, Kreiner says. Warrington and Kreiner have set this up so that a customer ships samples to the contract lab, PhyloTech gets the raw data back, analyzes it, stores it for two years with its cloud computing vendor, and serves up a secure web link to the customer. Clients also get a graphical representation that makes it easier for a biologist to interpret the data without help from a bioinformatics specialist.
Early on, the focus will be on serving academic customers. Part of the challenge will be in keeping the company focused on what those customers need. At the beginning, most of them want a comprehensive “multi-plexed” tool to look at the microbiome. But when I asked about more narrow uses, like testing for certain forms of E.coli that might contaminate a spinach field, Kreiner suggested that market might be better addressed by a lower-cost, customized chip. Such custom chips might be useful for specific human health applications, like spotting bacteria in the gut that could be related to, say, colorectal cancer, or chronic obstructive pulmonary disease. But those are markets PhyloTech won’t be pursuing from the very beginning.
“We almost have too many opportunities,” Kreiner says. “There are so many out there, we will need to prioritize.”
