What we can — and can’t — learn from our pets’ DNA

People are turning to DNA tests to learn more about the breed, behavior and health of their dogs and cats


This is Sweetie when she was a young dog. Is she part greyhound? Part Labrador? Her owner gave her a DNA test to find out. (A doggie friend digs in the background.)

L. Gunter; adapted by L. Steenblik Hwang

Sweetie, now 12, looks kind of like a greyhound. Or maybe a Labrador. She’s long and lean, with straight, silky fur, a happy-go-lucky face and floppy ears. Mostly, Sweetie looks like, well, a sweetie. She is a dog, after all.

Sweetie is 12 years old now. More than 95 percent of dogs in shelters in Arizona and California are like her, a mix of two or more different dog breeds.
L. Gunter

“When I first got her, I was convinced she was a labradoodle reject,” says Lisa Gunter. Gunter is a psychologist — someone who studies the mind — at Arizona State University in Tempe. Her research focuses on how people perceive dog breeds. She couldn’t help bringing her research home to Sweetie.

Labradoodles are a mix of Labrador and poodle. When someone breeds a Labrador and poodle together, the puppies sometimes get a poodle’s curly coat — but not always. DNA is the long string of instructions that tells an organism’s cells what molecules to make. Maybe Sweetie just got the DNA for smooth hair instead of poodle curls.

Gunter adopted her dog from a shelter in San Francisco, Calif. She didn’t know what breeds Sweetie’s parents might have been. And Sweetie wasn’t telling. To find out, Gunter had her dog’s DNA tested with a kit from Wisdom Panel. This company provides the tests Gunter uses for her own research. She swabbed Sweetie’s mouth and mailed the sample to the company.

A few weeks later, Sweetie’s results were ready. To Gunter’s surprise, Sweetie didn’t have any poodle or Labrador — or greyhound. “She’s half Chesapeake Bay retriever, which is rare for central valley California,” Gunter says. Her dog also is part Staffordshire terrier, part German shepherd and part rottweiler.

Doggie looks can be deceiving.

DNA testing for people is very popular. But now we can also check out what genetic traits a fluffy feline or pettable pooch carries in its DNA. We can learn what breeds a pet descends from, or in what region of the world its ancestors evolved. We can even try to predict how a pet might behave or what diseases it might face some genetic risk of developing.

But for all that these tests might provide some interesting results, they need to be taken with caution. Pet DNA tests aren’t necessarily as accurate as the human variety. And DNA itself isn’t destiny. Scientists and veterinarians are concerned that as DNA testing becomes more popular, people might confuse a DNA-based risk with illness — whether or not the pet is actually sick.

Playful pup or fraidy-cat?

The DNA in a dog or cat (or human!) comes in long, coiled strands called chromosomes. A dog has 39 pairs of chromosomes, and a cat has 19 pairs (humans have 23 pairs). These chromosomes are long chains of four smaller molecules called nucleotides (NU-klee-oh-tydz). The nucleotides occur over and over again — billions of times — forming long sequences. The sequence of those different nucleotides encodes instructions for cells.

Determining the sequence — or sequencing — those nucleotides was once a long, expensive process. So scientists came up with other ways to look at genetic differences between one individual and another. One of these depends on the fact that much of the strings of nucleotides, called sequences, are the same from one dog or cat to another dog or cat. (One cat may have stripes and the other spots, but both need the same basic DNA that tells cells how to, say, build a strand of fur. That sequence will be the same.) But every now and then, one of the four nucleotide building blocks has randomly been substituted for another.

It’s like misspelling one word in a long sentence or paragraph. These spelling mistakes are known as SNPs (pronounced snips). That’s short for single nucleotide polymorphisms (Pah-lee-MOR-fizms). Sometimes, a “spelling” glitch doesn’t change much. But in other cases, one alteration could change the whole meaning of the passage. In genetics, that one SNP may change at least part of the function of some cells or tissues. It could change a cat’s coat from striped to solid. Another SNP might make a pet more or less likely to get a disease.   

Sweetie (left) has a “sister” Sonya (right). Gunter and her wife did not get Sonya’s DNA tested because Sonya is a border collie they obtained from a breeder — so they know all about her family tree.
L. Gunter

Many genetic tests for dogs and cats search for patterns of SNPs. Different groups of SNPs can determine a dog’s breed or a cat’s ancestry, and some are linked to certain diseases. But these tests only look at SNPs that scientists already know about. There are many other potential SNPs waiting to be found. DNA also contains large regions that can be copied over and over, or that can end up deleted entirely.

That’s why Elinor Karlsson didn’t want to stop with SNPs. She wanted to sequence the whole doggie genome — meaning every single gene — letter by letter. Karlsson is a geneticist at the University of Massachusetts Medical School in Worcester. She’s got a special interest in mutts like Sweetie. “Mutts are just cool. Nobody knows anything about them,” she says. “As a scientist one of the things most fun to do is … seeing how much [of what] people think about dogs holds up.”

Karlsson is especially interested in behaviors. Dog breeders and scientists don’t know very much about what genes make a dog anxious or sad.

“Dogs and humans aren’t that different,” she says. “We study genetics to try and understand what makes people suffer from certain diseases, like psychiatric [Sy-kee-AT-rik] diseases.” These are disorders of the mind. “Dogs get psychiatric disorders,” she notes, much like people. They’re called behavioral disorders in pets. Dogs can suffer from anxiety, or become obsessive about chewing, retrieving or herding. Her laboratory has already identified a few candidate genes for obsessive-compulsive behavior in dogs. Her team published those findings back in 2014.

Sweetie and Sonya have a cat in the house, too! This is Henry. Cats can get their DNA tested, but most cats are not mixes of specific breeds, so they don’t have family trees that are as diverse as dogs.
L. Gunter

But getting enough DNA to determine dog behavior is a tough task. A curly coat or pointy ears might be controlled by one or a few genes. Behavior is much more difficult to pin down. One behavior could be controlled by many, many genes. To find them all, a researcher would have to study the DNA of thousands or tens of thousands of dogs, Karlsson says. “We couldn’t have a lab with thousands of dogs. It’d be extremely loud.” 

To get the DNA from so many dogs, Karlsson founded Darwin’s Ark. Like Wisdom Panel, Darwin’s Ark offers genetic testing for your pet. Karlsson’s test sequences every gene, not just SNPs. But it’s not quite as thorough as some human tests.

Sequencing every letter of the genome is a tricky process, like typing out a book as you read it. You’re bound to make a few spelling mistakes or miss some words. To address this problem, human DNA tests tend to run an analysis 30 times to fill in all the gaps. Write out the same book 30 times over and compare all the versions together, and you’ll end up much closer to the original.

Karlsson’s test on dogs tends to run through the genes just once. So there might be tiny regions that get missed. To make up for that, Karlsson adds more dogs. They will all have very similar DNA — they’re all dogs. And by sequencing enough of them, Karlsson hopes to fill in the DNA details that might get missed in only one sequence.

Looking for clues to attitudes

To learn about how a dog behaves, researchers need to survey its owners. Darwin’s Ark does this through citizen science — research in which non-scientists can take part. Pet owners fill out several long surveys giving details about their dogs’ personality. What do they like? What are they afraid of? By pulling such details from the surveys, Karlsson is hoping to match genes to a dog’s behavior.

That’s important, because people assume a lot about a dog’s behavior when they look at its breed. But maybe they shouldn’t, especially if it’s a mutt.

Sweetie, for example, has good doggie friends — but she’s not very good at making new ones. “It could be attributed to her American Staffordshire terrier or German shepherd ancestry,” Gunter says. When Sweetie loves someone, though, she is a real cuddle bug. Gunter thinks that could be due to those first two breeds. Or maybe it’s due to her Chesapeake Bay retriever or rottweiler traits. “You could tell a pretty compelling story with any of the breeds in her heritage,” she notes.

These are the breed results that Gunter got for Sweetie. There’s no greyhound or lab to be seen. Instead, Sweetie has one parent who was a Chesapeake Bay retriever, and another that was part German shepherd, part rottweiler and part Staffordshire terrier. See larger version.
L. Gunter

Scientists don’t yet know precisely how the behaviors of different breeds combine in a dog, Gunter points out. “Genetic influences of multiple breeds do not combine like dabs of differently colored paints or dashes of our favorite attributes,” she says. “I’m uncertain how informative it is to know the breed heritage of your mixed breed dog if we don’t know how multiple breeds affect behavior.” Maybe it’s better, she says, to just take your dog’s behaviors and work with them.

Adam Boyko is a geneticist at Cornell University in Ithaca, N.Y. He’s also the scientist behind EmBark, another dog-genetics test. He says some people learn the breed of the mutt and see a totally new dog. “We see a ton of owners that are so thankful to [learn] the breed mix because now they realize they have a better understanding of a dog’s behavior and things they can do to keep their dog happy,” he says. “They might find out their dog is part border collie and teach it to herd.” That might help it release some of its pent-up energy. Knowing what breeds are in their dog’s ancestry didn’t change the way the dog behaved. But it did change how people reacted to that behavior.

From DNA to disease

The DNA test that Gunter gave Sweetie didn’t tell her anything about Sweetie’s health. But some tests, such as EmBark, can do that. “What we can tell the owner is whether or not the dog has specific known genetic variants that are associated with certain diseases,” Boyko says. EmBark offers a test for more than 170 health conditions. These include ones where a DNA tweak may underlie some disease. An updated version of Wisdom Panel (not the one Sweetie got) offers a health test for more than 150 dog diseases as well.

Boyko’s lab has identified DNA tweaks that are associated with risks of seizures, heart disease and more. These data are of interest to dog owners. But they can be very important for dog breeders, Boyko says. These people want to know if a dog they want to breed carries genes that might boost a risk of certain diseases in its offspring. If so, maybe they would want to breed it with some other dog, or not breed it at all.

People love the squished-looking faces of pugs. But too much inbreeding means these animals can have difficulty breathing. DNA tests can help breeders know which animals should be mated together to make more pugs.
nimis69/iStock/Getty Images Plus

Cat breeders also want to know if their chosen breed carries the risk of some genetic disease. Basepaws is a genetic test that can investigate that. Wisdom Panel and a company called Optimal Selection also offer tests targeted to cat breeders.

Breeders and veterinarians can also send samples from their cats to a veterinary genetics lab at the University of California, Davis or to the one in which Leslie Lyons works. (Yes, that’s pronounced “lions,” and yes, she says, it’s very ironic.) She’s at the University of Missouri in Columbia. Lyons’ lab specializes in finding genetic links to diseases in cats. “The end goal for me is to improve the health of domestic cats. And one way to do that is to eradicate genetic disease,” she says.

But her hopes go far beyond felines. “Ultimately, we’d like to say this cat disease models that human disease or dog disease,” she says. If certain treatments for that disease work in other species, she notes, “we can apply them to cats.” And her findings might work the other way around, too. A treatment that works in a cat might later be tried in dogs or people.

Oscar is an orange tabby cat, classified as a domestic short hair. He doesn’t belong to any specific breed.
S. Zielinski

Unfortunately, people sometimes take these genetic tests as doggie dogma — that they determine a pet’s future health. In fact, they don’t. Even veterinarians don’t always know how to interpret the results of genetic tests for pets.

“[DNA tests] aren’t like other kinds of blood tests a vet does,” notes Lisa Moses. She’s a veterinarian at the MSPCA Angell Animal Medical Center in Boston, Mass. She’s also a bioethicist — someone who studies codes of conduct in medicine — at Harvard University in Cambridge, Mass.

Moses first heard about the DNA tests that people can get, such as 23andMe. The tests work just like Wisdom Panel and other dog-genetics tests. And people often misinterpret their results, she’s found. In fact, Moses didn’t know how to interpret them at first. “I just assumed if you had a positive [genetic] test, you had the disease,” Moses says. “And I think that’s what most people think.”

But that’s not true. Certain SNPs, deleted DNA sections or extra copies of some sequences are common in large populations. And some people who have them do indeed develop the illness they’re associated with. Yet most people who have them never get sick because of those genes, she notes. The same goes for dogs and cats.

Decode DNA with caution

Worries about genetic misconceptions keep bioethicists like Moses and scientists like Karlsson up at night.

After Karlsson had published papers on dog genetics, she began talking to people from companies that test dog DNA. She suddenly realized that “people could just start offering tests [based on] my papers.” This horrified Karlsson because she knew a single research paper is only the beginning of understanding what a gene variant might do. Many more studies would need to be done before she could firmly link a gene variant with some disease.

How reliable are different dog DNA tests? C&EN Speaking of Chemistry tested their resident pup, Ultraviolet, to find out.
C&EN/ACS Productions

“I knew those results weren’t good enough for a genetic test,” she says. “But there was no regulation that would stop that from happening.” There is no government group to decide or rule whether a dog- or cat-DNA test is a good one or not.

Horrified, Moses and Karlsson got together with their colleague Steve Niemi. He’s a veterinarian and director of the Office of Animal Resources at Harvard. They published an article in Nature on July 26, 2018. It pointed out that many of the genes that companies interpret as a test for diseases in dogs might not stand up to follow-up studies. The report also noted that tests of human and pet DNA can make mistakes.

The paper begged companies that test a pet’s DNA to set strong standards for which genetic sequences and diseases they attempt to link, and how they interpret the findings for breeders and pet owners.  

Boyko also says people should be careful when making decisions about vet care based on a DNA test. A DNA test can only offer warnings of risks. A dog that has a gene associated with blindness is at risk for blindness, he notes. But it’s not necessarily blind. “What we’re telling the owner is what you need to look out for,” he says. The next stop should be a vet who can monitor and test your animal now and in the future. The DNA results will be helpful there, Boyko says, because the vet will have a better idea of what tests to run.

And then a person would have to decide whether or not to run those tests. A human can know their dog has a DNA-based risk for a disease. But the dog doesn’t know the difference. Regular vet visits can be stressful for some dogs, notes Moses. Pets have different needs than people do. And in some cases, it might be easier on a dog or cat to not run the tests. In other cases, the test might be just fine.

In the end, your cat or dog is still your pet. “We want explanations; those are satisfying,” Gunter says. “We want to understand what makes our dogs who they are. But in a lot of ways we know that, we know who our dogs are.” Our pets are more than their DNA and breed and background. They are our companions and friends. We don’t need to know their DNA to know who they are. We just need to pay attention.

Sweetie didn’t become more terrier-like when Gunter read her DNA results. Her personality didn’t change when Gunter learned about her background. Those DNA results added to what Gunter knew about her life story. But the DNA test didn’t change the dog. Sweetie, in the end, is still Sweetie.

Bethany Brookshire was a longtime staff writer at Science News Explores and is the author of the book Pests: How Humans Create Animal Villains. She has a Ph.D. in physiology and pharmacology and likes to write about neuroscience, biology, climate and more. She thinks Porgs are an invasive species.

More Stories from Science News Explores on Genetics