scorecardresearch Skip to main content

I used my mutt as a DNA-test guinea pig

Louie, who came from a shelter in Rhode Island, was a guinea pig for his owner as he compared pet DNA testing services.Scott Kirsner for the Boston Globe

My wife was the first to cave and buy a DNA test to figure out what kind of dog we’d adopted.

Louie came from a shelter in Rhode Island, and every time we’d bring him to the neighborhood dog park, people would ask — or hypothesize — about what breed the little white fluffer might be. Havanese? Malti-poo? Bichon Frise?

“No idea,” we’d say. “He’s a mutt.”

The first canine genome was decoded in 2005, led by researchers at the Broad Institute in Cambridge. (They read the DNA strands of a female boxer named Tasha.) In 2009, Wisdom Health, part of the food and pet care giant Mars Inc., launched the first dog DNA tests for consumers.


A descendant of that product, the Wisdom Panel 4.0, was the one my wife bought, for $85. You collect a saliva sample from the pet, ship it to Wisdom, and wait two to three weeks to get the results by e-mail.

The verdict on Louie (above), according to Wisdom? 62.5 percent Chihuahua. The remaining 37.5 percent was an even split between miniature poodle, cocker spaniel, and English springer spaniel, Wisdom’s website informed us.

After that, when the dog was getting a bath, we might squint a bit and say, “Yeah, his eyes are kind of bulge-y, like a Chihuahua’s.”

The Wisdom test predicted that Louie’s adult weight would be 13 to 24 pounds. Currently, he clocks in at a little more than 14 pounds. Not bad.

Then, earlier this year, I bumped into Zenobia Moochhala, who mentioned that she’d recently taken a job as the chief operating officer of a Boston startup called Embark Veterinary. Moochhala was a cofounder of, the Waltham company that operates a marketplace for nannies and other caregivers.

My first question was, how was Embark different from Wisdom? More accurate, Moochhala told me. While Wisdom analyzes about 1,800 genetic markers, or DNA sequences that we know something about, Embark says it looks at more than 230,000. The company raised $10 million from investors in April, and it pays to be the official dog DNA test of the annual Westminster Kennel Club Show.


A while after our conversation, Embark sent me a complimentary test kit (it retails for $199), and I was again collecting spit from Louie’s mouth. His Chihuahua eyes bulged at me skeptically. “This time, it’s for work,” I explained.

After the sample was sent, I soon started receiving regular updates from Embark via e-mail. The sample was en route to the lab. The sample had been received. DNA processing was about to begin.

In mid-September, the results came in. Embark and Wisdom agreed about some aspects of Louie’s heritage, like ancestors that were probably miniature poodles and cocker spaniels. But Embark pegged him at just about one-third Chihuahua, and 36 percent rat terrier. Rat terriers, I learned from the site, were often kept as farm dogs in the early 20th century, “bred for catching barn rats in haystacks.”

What was I supposed to tell strangers? Did I have a dog that was mostly rat terrier? Mostly Chihuahua?

It was time to consult the experts. First, I rang up Elinor Karlsson, director of the vertebrate genomics group at the Broad Institute. She told me that “we don’t have a way, or haven’t agreed on a way, to measure the accuracy of the different tests.”


And if the providers of DNA testing haven’t (yet) published scientific papers about their analytical processes, “you have to take it with a grain of salt,” Karlsson says. “It could be 100 percent right, or not, and you’d have no way of knowing it.” (Karlsson confesses that she is a cat person: “Dogs are a fantastic genetic model,” she says. “But a cat’s purr — you can’t really beat that.”)

Next, I consulted Jerold Bell, an adjunct professor of genetics at Tufts University’s Cummings School of Veterinary Medicine.

“A mixed-breed dog that has been a mixed-breed dog for generations doesn’t necessarily have pedigreed parents behind it,” Bell says. For that reason, the testing could just say that a portion of a dog’s heritage is unknown — but “the consumer wants them to make a call on it,” he says.

Even Moochhala at Embark acknowledges that dog DNA testing is “a new enough space” that there isn’t yet an objective third-party authority that assesses whose results are the most accurate: “The answer is no, or potentially, not yet,” she says.

The other aspect of canine DNA testing is about health issues. Embark’s test told me that Louie was free from 14 genetic conditions that were “common in his breed mix” — everything from eye problems to seizure disorders to exercise-induced collapse. There were also more than 100 other health conditions that the test said he is free from, but many of them only afflict specific breeds.

That was good news . . .


. . . Until I talked to Karlsson at the Broad, who notes that “there is only a tiny percent of diseases, many very rare, that we can test for.”

She and Bell told me that it isn’t yet possible to flag dogs that are likely to develop conditions such as heart disease, diabetes, or cancer. Karlsson says that in her conversations with pet owners, when tests came back clear, as mine did, “I discovered that people thought their dog wasn’t going to get sick, that they were genetically fine.”

Not the case. And Bell worries about DNA tests raising alerts on some serious health conditions, such as a spinal cord disease called degenerative myelopathy, that will cause distress for pet owners — “but in most dogs, [having the gene] does not cause the disease at all,” he says.

Bell is a practicing veterinarian in Enfield, Conn., so I asked him what he tells his clients about DNA testing. “There’s still a large novelty component to it,” he says.

As people in the United States spend increasingly more on their pets — about $75 billion this year, according to the American Pet Products Association, a trade group — the dog DNA testing market seems likely to keep growing. Embark now employs 58 people, Moochhala says, up from 35 at the start of the year.

What am I telling people at the dog park about Louie’s background?

“It’s complicated. He may be mostly Chihuahua, or mostly rat terrier. But if you get a dog DNA test done, I’d recommend you do just one.”


Scott Kirsner can be reached at Follow him on Twitter @ScottKirsner.