OpGen's public debut was hardly auspicious. One of the exhibitors at Marco Island's Advances in Genome Biology and Technology meeting in February, the company's booth was all the way at the back of an aisle, behind a partition. Its biggest advantage was being next to the empty trays where attendees stacked their used plates and glasses after a coffee break.
But for the folks at tiny OpGen, a Madison, Wis.-based genome mapping company, the emergence from stealth mode at Marco Island was a smashing success. In the months since, the company has recruited more staff, exhibited at two other shows, signed its first commercial license, and assembled a scientific advisory board boasting Kary Mullis, Michael Waterman, and UCLA geneticist Julie Korenberg.
Not bad for a genomics startup in this economy. But then, OpGen isn't your typical startup — it's the antithesis of so many of its fellow companies, which tend to take a great tech concept and build a company while simultaneously developing and proving that concept. OpGen has only been around for two years, but it's founded on David Schwartz's optical mapping technology, which spent 14 years in the making. If OpGen is as successful as its execs, investors, and advisors predict, it could be a new model for business in the genomics market.
With applications in biodefense, pharmacogenomics, comparative genomics, and sequence assembly, optical mapping could grow exponentially as it stakes its claim to customers in the sector. "DNA would have to disappear from the face of the planet for this to become useless," says Colin Dykes, OpGen CSO.
Developed by the University of Wisconsin's Schwartz, known through the field as the inventor of pulsed-field gel electrophoresis, optical mapping is a whole-genome mapping technology that relies on restriction enzymes rather than probes. Starting by stretching out and electrostatically immobilizing strands of DNA on a glass surface, a six-base-cutter enzyme is sent in to snip apart the DNA strands at matching sites. The DNA then shrinks back at the edges of each cut, and the map is made by visually identifying the position of all resulting gaps. That map can be used to build a sequence — by assembling reads based on the location and spacing of the six-base matches — to identify strains of an organism, compare isolates, or find telltale indicators to predict patient response to therapeutics.
When Schwartz first started working on this technology, he wasn't sure how it would be used — his own goal was to study centromeric structure. Today, Schwartz and his OpGen colleagues say the technology is becoming increasingly relevant. OpGen investor and acting CEO Dan Broderick points to papers published earlier this year indicating that genetic diversity is less a result of SNPs than of insertions and deletions. Optical mapping, which is not a SNP discovery tool, lends itself to finding genetic rearrangements.
As optical mapping finds its footing in genomics research, analyst Jim Golden says that the technology and the company will have the first impact on individual whole-genome analysis, long before sequencing companies such as Solexa or 454 can make a dent. "I truly think OpGen will be the first company to be able to do the whole human genome and give you [medically useful data]," he says. Golden, who first studied optical mapping as a competitor when he was business development manager at 454, calls it "elegant, simple, inexpensive, and useful."
PCR inventor Kary Mullis, who sits on few advisory boards, was eager to sign up for OpGen's. Schwartz is a "really brilliant person," he says, adding that for applications seeking to uncover large-scale features particularly between human genomes, "I don't know of any other way [besides optical mapping] to get at that information."
David Schwartz was at a 1982 Cold Spring Harbor Laboratory meeting when he first saw a movie of single DNA molecules. It was the start of what has since become his love for manipulating single molecules. He abandoned his pulsed-field electrophoresis work to study them and, by 1989, wound up with preliminary data for the earliest phase of a restriction map using optical mapping.
Four years later, Schwartz, then in the chemistry department at New York University, published his first paper on gel-based optical mapping (an approach that later gave way to mapping on a glass substrate).
By 1995 and after several more successful papers, Schwartz was thinking of starting a company to commercialize the technology. But Ciba, now part of Novartis, and Chiron Therapeutics partnered to offer a three-year, $10 million-plus licensing deal to Schwartz's lab to automate the platform and make it more robust. The generous funding meant it made more sense to stay in academia, Schwartz says. "Instantly we could move forward and I wasn't burdened with the issues of starting a company."
With the funding flood, Schwartz's NYU lab swelled to 45 people and the technology was propelled forward — automated, operating on a glass surface, and using genomic DNA. The goal at the time was studying centromeres, notoriously repetitive regions that optical mapping could finally make sense of. But an early grant application to NIH's new genome study section showed that "they weren't necessarily interested in centromeres," Schwartz recalls. So he shifted the focus to whole-genome analysis and worked on refining chemistries, developing an image acquisition system using high-powered microscopy, and building map assemblers.
By 1998, the private-sector funding ended, but Schwartz's lab was able to make up similar amounts through public agencies — NIH, Burroughs Wellcome, DOE, and NSF continue to fund his lab today.
At the same time, Schwartz was already looking at avenues other than NYU for his lab. "I felt that the future of biology was going to lie at the intersection of biological scientists and [engineers]," he says. NYU's lack of an engineering school was a weakness, and he headed for greener pastures in Wisconsin, which he chose in part for its renowned tech transfer agency, the Wisconsin Alumni Research Foundation. Schwartz was ready to saddle up and think about starting his own company again
Incorporate or bust
Despite the maturity of the technology, there was no easy path to incorporation. Schwartz called his friend Stan Rose, who had just sold off his array company Genetic MicroSystems to Affymetrix, and asked if he had any time for another startup.
Rose, now chairman of OpGen, calls Schwartz "one of the few creative geniuses working in molecular biology today." He saw optical mapping as an "opportunity to have a big impact on science and healthcare" and agreed to help get the company going.
The next step was finding backers. WARF facilitated a meeting with local VC Dan Broderick from Mason Wells. Boderick was immediately impressed with the technology and the people: direct analysis of single molecules "took a lot of error out of the process of analyzing DNA," he says. And Schwartz won him over, too. "I was investing as much in Stan and David as I was in what I understood about the technology." By then, Broderick adds, Schwartz had some 15 peer-reviewed papers published on the technology — rare validation in a field of unproven startups. A tech patent was also on its way to issuing.
Broderick's first offer for a half-million-dollar seed round came in the summer of 2000, but "we couldn't come to terms on a valuation," he says. A year passed. Broderick heard that Schwartz had never gotten funding, and offered again; the deal went through. The seed money took care of writing a business plan and testing the technology's throughput. In November 2002, the series B round closed with $2.3 million, which continues to fuel the company. The first step was changing the name: OpGen started out as eDNA, which, Broderick likes to point out, some word processors automatically change to "Edna."
Golden sees it as a huge advantage that the company's gotten so far on so little money. It makes OpGen more likely to reach a profit sooner, for one thing. And "if you go out and raise $50 million like US Genomics did, you'd better IPO at $400 million or your investors are not going to be happy," he adds.
OpGen is currently fundraising for its third round. "It's going as well as can be expected in today's marketplace," Broderick says. The goal is to secure $7 million to $10 million, although backup plans provide for getting as little as $5 million. Completed genome maps of organisms such as Deinococcus radiodurans and Plasmodium falciparum and soon-to-be published ones of Trypanosoma brucei and Leishmania major will likely help the cause.
Progress and problems
Innovation continues for the technology, both at Schwartz's academic lab and at OpGen. Many, but not all, advances are shared between teams. Dykes says that OpGen, which operates on a service model — customers ship samples to Wisconsin and pay on a per-megabase scale, although each customer negotiates an individual contract — is building its own computer cluster. When finished, the cluster is expected to be "in the top 120" for computational power, Dykes says.
The service model is one challenge for the company, Dykes acknowledges: "The last two people we spoke to said they wanted to buy systems." But the computationally intensive analysis makes the system a cumbersome thing to put in a box.
Another challenge, oddly enough, is the very maturity of the technology. Dykes calls it a "mixed blessing" — obviously, OpGen has a tremendous advantage thanks to how proven the technology is, he says. "The downside is that when people think of it, they think of it as it was described maybe two or three years ago." One human genome map, which originally took five years, now comes together in five days — and Dykes hopes to get that down to five hours within a year.
Golden believes the complexity of the technology is also a hurdle: "getting people to understand the value" will prove an undertaking, he predicts. But if OpGen can overcome that, he adds, "I think these guys actually do have the killer app."
Bringing that app to market will be the next step. "We hope that the first impact will be in cancer clinical trials, predicting responders versus nonresponders," Broderick says. After that, cardiovascular disease or CNS could follow.
The potential for other markets will also be important, Schwartz hopes. "The thrust now that we're finally able to address is populations," he says, envisioning the mapping of loads of human genomes for solid comparison work. Also, the technology offers sequencing centers and others the "chance to survey the whole genomic landscape and then pick" which organisms are worth full-blown sequences, he says.
The possibilities, as Schwartz and the OpGen crew see them, are endless. And they're looking forward to the ride. "We intend to explode onto the scene in the next year," Dykes says.
During the course of 14 years working on optical mapping, David Schwartz has learned a thing or two about commercializing technology. His advice to fellow scientists looking to found a company:
Choose your partners wisely. "There are a lot of unexpected events that you're going to pass through that can stress friendships."
Roll with the punches. "Be prepared to change what you want to do in terms of the company. … Don't get too wedded to your own ideas."
Think about sticking to science. "If you're going to play businessman, make sure you are as good a businessman as you are a scientist. … The business aspect is just as important and probably even more important as the scientific aspects of the business. If you can't deal with that, don't do business."
The Golden Rules
Analyst Jim Golden evaluates companies based on five areas of risk. OpGen's standing, he says, makes it "a venture capitalist's dream."
Risk area Golden's evaluation
Technical "They've got it locked."
Managerial "They're doing fantastic."
Financial "They need to raise some more money, but they're doing fine."
Marketplace "There's definitely a market need. So many pharmas are looking [for this kind of technology]."
Competition "I don't see them having any competitors in the near term."