Two scientists walked into Leroy Hood‘s office in Seattle a little more than five years ago with a burning question about the genomes of dogs. They had sequenced the entire string of DNA in the canine genome for biomedical researchers, and sought out the high-speed gene sequencing pioneer for business advice. They wondered what it would take to build a business that sells information to pet owners, who might be curious about what breed their dog is.
This conversation in the summer of 2003 led to the formation of Seattle-based Argus Genetics, a spin-off company from the Fred Hutchinson Cancer Research Center. The technology has since been licensed to Mars Inc., the $21 billion-a-year candy and pet food giant in McLean, VA, which has turned it into a test called the Wisdom Panel MX. I heard the story of how this idea evolved into a product from Carl Weissman, a venture capitalist at Seattle-based Accelerator (and OVP Venture Partners), and Neale Fretwell, a business development manager for Mars’s veterinary division.
The original idea from Elaine Ostrander and Leonid Kruglyak at the “Hutch” intrigued Weissman when he got the referral, even if it made some of his investing partners roll their eyes because it wasn’t their usual swing-for-the-fences-to-treat-cancer kind of bet. Weissman and Kristina Burow, a Kauffman Fellow now with Arch Venture Partners, started their research. They learned there are 60 million dogs in the U.S., and half are purebred, while half are from mixed-breed heritage. They talked to breeders and purebred owners, and found out they didn’t want the test because it might disqualify purebreds from shows. That left mixed-breeds. There are 30 million of those dogs, and the test could be developed and sold for $100. That meant if Argus could capture at least 11 percent of the market, it could have a $330 million opportunity.
“Dog people are serious,” Weissman says. “It looked like an interesting business to us.”
It also seemed like something that could take off quickly, Weissman thought. It was an application of cutting-edge science that didn’t need to pass the strict scrutiny of the FDA, or go through hassles with reimbursement from health insurers. He negotiated for the rights to the technology from the Hutch, and set up the company in a structure that left what he calls “significant ownership stakes” for Ostrander, Krugylyak, and the Hutchinson Center, who get royalties on sales from Mars. (Ostrander has since left for the National Institutes of Health, and Krugylak is now at Princeton University.)
Mars got an early bead on this technology because it has sponsored some of Ostrander’s research in the past, Fretwell says.
Here’s how it works. A pet owner comes into a veterinarian’s office, curious about what breed their dog is. They provide a blood sample, which Mars returns to the vet within two weeks. The development of the test required an analysis of 19 million DNA markers from 13,000 dogs. The answer doesn’t come in precise percentages, but rather in ranges that would say a dog is more than 50 percent of a particular breed, like border collie, and gets between 25 and 50 percent of its heritage from another breed, and a smaller amount, say 12 to 25 percent, from another genetic line. The test is made to recognize 99 percent of American Kennel Club certified breeds. It’s considered 90 percent accurate—the best available on the market. Cost? $125 to $150.
Which sounded like a lot to me, if all you’re talking about is satisfying the owner’s curiosity. So I asked Mars’s Fretwell.
This type of test is unusual, but not unique. MMI Genomics of Davis, CA, markets a test for $120 that requires a cheek swab and no trip to the vet. That test isn’t as accurate, Fretwell says, because cheek swabs provide lower yields of DNA, and can get mixed up with food and bacteria. Mars’s strategy is to go through veterinarians instead of selling direct to the consumer, because vets can help owners interpret the information—like whether your dog is part Golden Retriever, and therefore needs more exercise than it’s getting to stay healthy. Sometimes the test can provide surprising results, showing a dog with a short black coat can actually be more than 50 percent Golden Retriever.
“We think vets are an important part of the process of helping people use the information,” Fretwell says.
So far, this test hasn’t really struck a chord in the marketplace. Mars, as a private company, doesn’t disclose its sales. The test has recently gotten some publicity on ABC’s “Good Morning America” and on “The Rachael Ray Show.” Fretwell would only say that the product was introduced nationally in February 2008, and it remains “in an early adoption phase.” Weissman was a little more candid, saying, “It hasn’t taken off yet the way we thought it would.”
I know one person who has had the test done—Spencer Lemons, a Seattle-based tech transfer consultant who used to run the licensing office at the Fred Hutchinson Cancer Research Center. He got his black-coated dog, “Bear,” from a rescue shelter in Bothell, WA. Lemons suspected it had some border collie and Australian shepherd heritage, but wasn’t really sure.
The Wisdom Panel test results showed “Bear” didn’t really have a dominant breed, although he was part Cocker Spaniel, part Field Spaniel, and part Chow Chow. I didn’t get the sense this bit of information changed Lemons’s view of his dog very much. He still sent me a picture of his companion, and made sure that I saw his “pretty blue eyes.”
(Correction: This story initially listed the accuracy rate for the Wisdom Panel at 84 percent. An upgraded version of the test introduced in October raised that level to 90 percent.)