What is DNA Testing for Dogs?

DNA testing is a rather new option (for humans and canines). In 2007, when the first dog DNA tests kits hit the market, many breeders, veterinarians and dog enthusiasts were excited. While some good has come from Dog DNA testing, there have also been a number of significant disappointments as we have realized their limitations.

What is DNA testing? You might remember from Junior High biology that DNA is “code” for deoxyribonucleic acid.

DNA is essentially a nucleotide sequence. Every living organism possesses DNA that functions as a one-of-a-kind blueprint with information about that human, plant, bacteria, cat, bird, or dog.

For our discussion, individual genes are one of the most important aspects of DNA. Specific Genes make up a dog’s genetic code. These genes can be coded for individual proteins which determine the building and reproduction of cells in a dog.

A human’s DNA is made up of 23 chromosome pairs (46 total chromosomes) with a total of about 3 billion DNA base pairs.

Canine DNA has similarities to human DNA, but also significant differences. A typical dog has 20,000 to 25,000 genes which are positioned along 78 total chromosomes. A dog’s DNA has about 2.5 billion nucleotides. Dog DNA testing focuses on a tiny percentage of those.

DNA's four letters (aka bases) are A, T, C, G: adenines (A) pairs with thymines (T), cytosines (C) with guanines (G)

For human DNA, the possible alphanumeric combination is 3.2 billion letters. Thus, each human has a unique DNA. In a similar way, a full sequence mapping of a dog’s DNA (first done by MIT researchers on Tasha the female boxer in 2005) gives a blueprint of that particular dog.

When should Dog DNA Testing Be Done?

DNA testing will be more accurate if it is done after a dog is at least two years old. Puppy traits are constantly changing. For instance a Poodle’s coat is typically not very curly as a puppy, but it continues to get curlier as the puppy ages. An early test is less likely to detect the curly coat if it has not yet manifested.

Wisdom Panel (one of the more popular K9 DNA tests), says on their website, “Note that it can be very difficult to observe breed traits in puppies, because they are growing and developing rapidly. Most dogs will not achieve their final mature physical traits until they are at least one to two years of age.”

Dog DNA Testing Process

If you want to have your dog’s DNA tested, the typical process would look like this:

1. Buy a DNA test packet online or from your vet. Some of the most popular DNA test kits are Embark, Wisdom Panel,
2. Collect the DNA sample. In the past this was done through a blood draw (by your veterinarian), but now the DNA is most often collected with a cheek swab.
3. Mail the DNA sample to the makers of the test kit.
4. At the Lab, your DNA test provider (Embark, PetConfirm, etc) should extract DNA from your dog’s cells in the sample swab (or blood draw) that you sent them.
5. DNA markers will be evaluated (depending on the test used, it will range from dozens to hundreds of markers that are evaluated).
6. Certain markers are analyzed (A, T, C, G). These markers give a key to your dog’s genetic code. Only a small percentage of your dog’s total markers will be analyzed (dozens or hundreds); Remember for humans there is a possible combination of 3.2 billion letters. For $85 to $200, it is not realistic to expect a full DNA mapping of your dog. The DNA test kit, will test a very small percentage of your dog’s overall DNA.
7. Your dog’s markers (eye color, coat curl, ear size) are compared with markers collected from registered purebreds in the past. The larger the collection from registered purebreds, the more accurate the comparison will be.
8. Most DNA testing services will consider multiple potential family trees. In other words, your dog has specific characteristics, and the DNA testing service is comparing other dogs (of known ancestry) which have similar characteristics. The computer can present dozens of different models of your dog’s possible ancestors.
9. The computer looks at these dozens of possibilities and scores which one it thinks it’s the most likely combination for your dog. This gives you a range of potential ancestors for your dogs, and essentially the DNA testers present you with their “best guess” as to what your dog’s hereditary looks like. It is important to remember that this is merely a best guess and far from 100% accurate, as the DNA testers are quick to acknowledge.
10. This “best guess” can vary between litter mates, and certainly between different DNA testers. It’s very helpful for a Heinz 57 dog to have some idea of what your dog’s genetic markers point to for possible ancestry.

Benefits of Canine DNA Testing

Canine DNA testing has some huge potential benefits.

1. Determining Parentage of an Individual Puppy. Most breeders are very familiar with the AKC’s DNA Parentage Evaluation Service. It is possible for more than one male to breed with a female, and the breeder may question which male is the dad of a particular puppy. We have friends who breed purebred Labrador Retrievers and found themselves in this exact predicament. If the genetic factors tested match the DNA of one father, better than the other, it becomes clear which Labrador was the father of the litter. AKC’s DNA Parentage Evaluation is a great use of DNA testing.
2. Linking an individual dog with its pet waste. There are currently efforts to identify which dog’s poop was left by an less-than-thoughtful dog owner going for a walk (PooPrints is the best example of this). In the same way that a genetic pattern of human DNA (blood, hair, skin cells) can often be used to match a murderer to the scene of the crime, a dog’s waste matter can be used to accurately match with the dog it came from.
3. Helping Breeders know what Puppies a Dog Will Produce. DNA tests are very helpful in letting a breeder know if a poodle carries the furnishing gene, whether a dog can produce the phantom markings in a puppy, the likelihood of certain coat colorings, etc.
4. Recognizing Potential Diseases. As researchers collect DNA information about genetic tendencies toward certain diseases, a DNA test can be helpful in knowing whether your dog is pre-disposed to a certain problem.
5. Giving the Best Guess of Potential Breed Origin to Someone who Adopts from the Humane Society. For decades, families who have adopted a rescue pup of completely unknown origin, have asked friends and veterinarians for their best guess as to what breeds make up the dog. DNA testing gives a more scientific means of arriving at a best guess, but the results are far from perfect (see below).

Problems with Dog DNA Genetic Testing

There are significant inabilities of Dog DNA testing. There is no regulating body for dog DNA tests (The FDA regulates the human DNA test, 23andMe, but nothing similar exists for dog DNA testing). Families who have sent DNA to multiple vendors, have received back widely varying results. And different results have been given with the same litter from the same canine parents and even with the same dog from the same DNA tester when samples were sent at different times.

Widely Varying Results of Dog DNA Tests

You can read humorous stories online about numerous people have sent saliva to multiple DNA testing groups and received widely varied results. Even the companies that do the DNA testing point out that they're not entirely accurate, but can give you a general idea of your dog's ancestry.

That's why purebred dogs (AKC registered) often will test with some "mixed breed" in them, and why litter mates will often test differently. Routinely dogs will get back 3 different sets results when sent to 3 different DNA testing groups. Also many mixed breeds of known ancestry (AKC papers on parents) come back from the tests as something different. In other words the DNA tests usually get you in a close ballpark (“best guess”), but are far from 100% accurate. 

The tests themselves point out that pups within a litter can vary. For instance, within a litter of f1b Goldendoodles (75% poodle and 25% Golden Retriever), we could test one of the much curlier siblings, and that sibling would be more likely to show the 75% poodle ancestry; whereas a less curly f1b Goldendoodle pup, might display less of its Poodle heritage in the DNA test. 

Although DNA tests seem to be improving, multiple DNA tests have been shown repeatedly to be unreliable to arrive at exact ancestry. Consumers who think that DNA tests are an exact science do not understand their true nature.

For example, the Wisdom Panel test acknowledges that purebred registered AKC dogs, may well show up with other things listed:

They say "the Wisdom Panel Canine DNA tests are not intended to refute" known ancestry and that "Because there is only one purebred tree model in the 11 considered, the statistical process inherently favors mixed ancestry." The inherent bias toward multiple breeds is helpful to someone who adopts a puppy from the Humane Society and wants to find out every possible breed their puppy might contain. Particularly if some breeds that may (or may not) be present or more prone to specific health issues.

Indisputably Wrong Results

A quick internet search will bring up examples of dog DNA tests that were so wrong, they’re laughable. People have sent the DNA from the same dog to the same testing center at different times and received back different results (https://news.vin.com/vinnews.aspx?articleId=23206). A veterinary group took a Purebred registered Poodle and submitted it to DNA testing groups as a Purebred, and the results came back that this was a purebred Poodle. When they submitted the same DNA to the same groups, but submitted the DNA as a mixed breed (selected that option), they were told that the exact same dog contained several other breeds in addition to standard Poodle. Supposedly, their AKC purebred Poodle was also part Labrador retriever, Pembroke Welsh corgi, Shih Tzu, Tibetan spaniel and Miniature Poodle.

Another example:

On a message board  of the Veterinary Information Network (VIN), an online community for the profession, veterinarians reported a variety of results in friends, clients’ and their own dogs. Some were plausible and some were wacky.

On the wacky side was an 80-pound dog whose mother was known to be a Labrador retriever that tested as mostly miniature poodle plus Yorkshire terrier.

There was a pedigreed dachshund who turned up as predominantly Siberian husky, with a dash of dachshund and Ibizan hound.

There was a dog that looked classically Chihuahua that came back as “an extremely complex mixed-breed dog ... (with) distant traces of Afghan hound, Cavalier King Charles spaniel and toy fox terrier.”

From Online Article: “Dog breed genetic tests put to the test: Science is solid but results aren’t precise”

Dog DNA tests often appear different for Sibling Dogs.

On their website Wisdom Panel addresses the question of why Siblings so often test different (it shows a completely different ancestry for two pups from the same parents).

Wisdom Panel writes, “Another factor involved in litter genetics is which exact segments of DNA each puppy inherited. While each parent is responsible for passing down 50% of their genetic makeup to a puppy, the actual composition of that genetic contribution varies between siblings. This is because of the random nature in which genes are passed down and a process known as genetic recombination. As we can see, in many litters, particularly those of mixed-breed dogs with a number of ancestral breeds, puppies can show a wide variation in their physical traits like coat color or length, leg length, and even head and ear shape and size. All this variation tells us that each puppy really is truly, genetically unique!”

Those familiar with genetic testing should not be surprised the canine siblings often test different as the same thing happens with humans. Human siblings with identical biological parents often find that a DNA tests tells them they are from different genetic backgrounds. Even DNA tested twins can be told they’re different percentages of various nationalities.

https://news.nationalgeographic.com/2018/03/dna-ancestry-test-siblings-different-results-genetics-science/

Elizabeth Warren’s attempt to prove that she had some Indian blood became a national news story and also showed the inability of DNA tests to perfectly determine human nationality in a satisfactory way.

https://www.factcheck.org/2018/10/the-facts-on-elizabeth-warrens-dna-test/

For an interesting study of two human siblings that had very different results, see the following article, which says only identical twins would have the same results for their ancestry:

https://homedna.com/blog/why-can-siblings-have-different-ancestry-results

Dog DNA Breed Testing is a Best Guess

One veterinary research firm took mixed breed dogs of known origin from the University of Pennsylvania’s School of Veterinary Medicine.  They sent DNA samples of these dogs (of undisputed origin) to Wisdom Panel and Canine Heritage.

For instance one of the dogs (Salem) was “exactly 68.4 percent beagle, 14.4 percent German shepherd, 9.4 percent giant schnauzer and 7.8 percent basset.” The tests that came back were disappointing. Wisdom Panel was the most accurate, even though it was still quite wrong.

Wisdom Panel found the beagle in Salem, along with a touch of German shepherd. But it very wrongly detected Ibizan hound in Salem’s background. This made the researcher (who owned Salem) chuckle as it was clearly an incorrect finding. None of Salem’s ancestors were anywhere close to an Ibizan Hound. 

The researchers helpfully concluded, “Salem’s result demonstrates that such charts [from Wisdeom Panel] should not be accepted as a precise family tree but merely a guideline when dealing with complex mixes. The chart shows one of her parents as purebred beagle. Casal said Salem's maternal grandmother was a purebred beagle but not her parents.”

Wisdom Panel had incorrect findings, but they were in the ballpark. The Canine Heritage test was far worse, not even recognizing the Beagle. The Canine Heritage test could not find any primary breeds. It named German Shepherd as a secondary breed, and Chesapeake Bay retriever as “in the mix.” 

The important thing to remember is that Dog DNA Breed Testing is a best guess as to what dog ancestors may have been.

If you are still interested in further research, please read the articles below to read about mixed results in DNA testing, as a reminder that this is a developing science that is still just a best guess.

https://www.whole-dog-journal.com/issues/21_7/features/Dog-DNA-Tests-Mixed-Results_21872-1.html

https://www.wfla.com/8-on-your-side/dog-dna-testing-the-test-kits-reveal-startling-different-results/1494814305

https://help.wisdompanel.com/s/article/I-tested-siblings-and-the-results-are-different-how-can-that-be

https://help.wisdompanel.com/s/article/Wisdom-Panel-test-said-my-registered-purebred-dog-isn-t-a-purebred-How-can-this-be

http://fortune.com/2015/06/25/dog-dna-tests/

Potential Dangers of Dog DNA Testing

Typically veterinarians are not genetic experts, and the DNA tests can push them to make decisions about individual dogs without factual results to back it up. A vet going on a widely divergent “best guess” from someone’s DNA kit, might feel pressured to put a dog down even when he/she sees no specific physical reason to do so.

On November 12, 2018, In the Science section of The Atlantic, Sarah Zhang takes a hard look at problems with DNA testing (“What Vets Think of 23andMe for Dogs”). For example, she writes,

Several dog owners told me their vets were curious when they brought in a DNA report for their dog. When I reached out to the American Veterinary Medical Association, the American Animal Hospital Association, and several state veterinary associations though, most either did not respond or responded to say direct-to-consumer DNA tests weren’t on their radar. Meanwhile, the veterinarians and canine geneticists who did want to talk were largely skeptical.

“Veterinarians, we’re not really educated in clinical genetics, because it’s a brand-new field,” says Lisa Moses, a veterinarian in Boston. Moses was especially concerned about the health-risk information. Doctors can refer human patients to genetic counselors, she points out, but veterinarians don’t have genetic counselors on call for dogs. Moses co-authored a comment inNature earlier this year, in which she recounted the story of a 13-year-old dog that was losing her ability to walk. Her owners decided to buy a $65 direct-to-consumer test, which showed a mutation linked to a neural disease called degenerative myelopathy (DM). Convinced that she would slowly die of the disease, her owners put her to sleep.

But the mutation for DM is notoriously hard to interpret. Kari Ekenstedt, a professor of anatomy and genetics at Purdue University, calls it the “ever controversial DM mutation.” The problem, she says, is that not having such a mutation is a good sign a dog does not have DM, but having a mutation does not guarantee the dog has the disease. It’s possible the dog Moses wrote about had an entirely treatable spinal disorder and did not need to be put down.

The episode prompted Moses to take a closer look at the pet-DNA industry, and she came away even less certain of how to interpret the results. While the Food and Drug Administration regulates 23andMe, no one is looking at pet-DNA tests. Moses says she had been taking the DNA tests at face value, and she began to wonder if she had caused her patients too much worry by doing so. “I didn’t understand how iffy, how little there was for me to really take stock of these tests,” she said. Carrie Waters, a veterinarian in Dallas, echoed the sentiment. “There’s a number of labs doing it, but I’m not totally convinced there’s the best quality,” she said.

Statement from Wisdom Panel

Over the past many years, several puppies from Crockett Doodles have had DNA testing done. It has been interesting to see the results. So far (to our knowledge), none of the pups have tested as susceptible to any particular genetic disorder (for which we are thankful). The majority of tested pups, test exactly correctly according to their parentage, but we have had a handful of cases with slight variation. In these cases, the DNA tests were nearly accurate in their “best guess,” but missed it slightly. Some of the DNA tests misrepresent the test to the end customer making it sound like absolute scientific proof which can be confusing to their customer, who may actually doubt the known heritage of a doodle puppy. Below is an email exchange with Wisdom Panel after a particularly interesting DNA test.

Letter to Wisdom Panel about Incorrect DNA Results
Dear Wisdom Panel:

In the past, I've believed that you were one of the best DNA tests, but it is becoming more and more difficult to recommend you to anyone who gets a pup from us. I think you need to be more up front with those who purchase the Wisdom Panel test that it is the Best Guess as to a dog's heritage and that it is not fool proof.

Most recently a family got a mini Goldendoodle from our home-raised network (Crockett Doodles). Their female mini Goldendoodle (Poppy) was the offspring of a purebred registered mini Poodle (Rudy) and a purebred registered Golden Retriever (Heather).

Because we artificially inseminated, and Heather was not around other dogs, we can know for 100% who the parents were.

As is very typical in crossbreeds (especially first generations), some pups got more of the Golden Retriever and others received more of the mini Poodle. For instance some pups had a far curlier coat--or were larger or smaller.

Poppy received far more of the Golden Retriever's characteristics. She sheds like a Golden and has an incredibly friendly attitude like a Golden. Basically Poppy's adoptive parents told me that she is like a mini Golden Retriever (she got her size from her mini Poodle dad).

Some of her litter mates received more of the mini Poodle traits and would undoubtedly test 50% or higher mini Poodle. I recognize that in your FAQ section of your website you acknowledge that DNA tests often differ between siblings, should not contradict known heritage, etc. However the test results do not make this nearly clear enough and are VERY misleading to families who have not thoroughly studied genetics.

In this case, I thought your DNA tests were quite accurate (considering they test for 400 breeds). They identified 50% Golden Retriever (exactly right). But for the other parent, they predicted part mini Poodle (correct) part Cavalier, part Cocker (both incorrect).

This makes sense to me, because basically Poppy looks to me, like she received about 85 % of the Golden Retriever traits and 15 % of the mini Poodle traits. However, based on her smaller size, I can see where the test would recognize that she's not nearly big enough to be 85 % Golden Retriever, so in my mind the Cocker / Cavalier guess makes sense.

What is extremely bothersome to me is that your test results (the family kindly sent them to me) are incredibly misleading in making it sound like this is a perfect 100% Science, actually making the adoptive family question the known factors. We know that the dad was a registered mini Poodle and the mom a registered Golden Retriever. We do not have a single mini Poodle / Cocker / Cavalier male in our network--anywhere, so clearly this is not correct. But your test presented it as if it was perfect Science. This is absolutely deceptive advertising.

You're probably aware that the human DNA testers are embattled with class action lawsuits. I imagine that pet DNA testers will not be far behind, unless you more clearly specify that this is your best guess and should not be used to contradict known realities.

For instance Poppy's family initially wanted a full refund, since their pup did not test 50% Golden Retriever and 50% mini Poodle. I talked to them on the phone and they're very reasonable, and as they study it more and think through the realities (just because a mom is a Golden Retriever and a dad is the mini Poodle, does not mean the pup will get 50% of one parent and 50% of the other), they're understanding it better.

Again, for what it does (narrow down the top possibilities out of 400 breeds), the Wisdom Panel is remarkable. But to present it as something far more than it actually is becomes misleading to your customers and increasingly frustrating to reputable breeders.

Can you please send me something in writing acknowledging that your test is not 100% accurate, and that it should not be used to try to contradict known results of a puppy of known heritage?

Thank you; I'm not trying to come across as angry, I'm just frustrated that the marketing of your product seems to assume a level of certainty that is not accurate.

Examples of recent lawsuits over wrong genetic testing:
https://topclassactions.com/lawsuit-settlements/closed-settlements/820889-23andme-dna-testing-kit-class-action-settlement/
<https://www.theatlantic.com/science/archive/2018/09/dna-test-race-lawsuit/570250/>

Wisdom Panel’s Helpful Response

I was very appreciative of the response from Wisdom Panel which acknowledged my concerns. I just wish the info given to their customers made this more clear from the start. They attached the PDF below to the following email.

Thank you for reaching out; we are sorry to hear that you are having issues concerning results that we may have provided. We do our best to be clear regarding what the Wisdom Panel Canine Breed tests are intended for; specifically posting several knowledge articles on this subject, as well as accuracy, and how the testing works on our help site (I have included some of them below), on the product description pages for our secondary online retailers, and within the online version of the report.

We caution our customers that visual identification using physical and behavior traits is often incorrect (studies have shown that even dog experts are incorrect about 75% of the time). When two dogs of the same breed are bred together there are a limited amount of genetic possibilities, and you can be reasonably certain of what the resulting puppies will look like. However when you begin to mix breeds together, even F1 crosses, additional genetic possibilities are opened up. The physical traits that are expressed by these dogs can often mimic those of other breeds, or make them look different than expected based on their ancestry. However physical appearance does not alter genetic breed ancestry.

The exact breed ancestry of puppies in the same litter depends on several factors. Each puppy in a litter is created by a unique egg and sperm, much like fraternal twins. Identical twins have not been shown to occur in dogs at this time. Because of this, we have found that average full siblings share approximately 2/3 of their tested DNA signature, that is, when they have the same mother and father. Because of the unique reproductive cycle of the female dog, multiple sires for the same litter are common. If a litter was sired by more than one male, each puppy could have a different father, thus making them half-siblings. The genetic relatedness of half-siblings is even lower, and can produce very different breed ancestry results.

Another factor involved in litter genetics is which exact pieces of DNA each puppy inherited. While each parent is responsible for passing down half of a puppy's genetic makeup, the actual composition of that genetic contribution varies between siblings, and is also influenced by the diversity of the parents. The semi-random nature by which genes are passed down is caused by the process known as genetic recombination. In genetic recombination, prior to production of sperm and ova, the parent's chromosome pairs swap pieces of DNA, then divide in half, resulting in a unique combination of DNA for each ova and sperm. This process happens with each litter, producing a tremendous amount of genetic "shuffling" over even a few short generations. On average, through the DNA of their parents, puppies receive about 25% of their total genetic makeup from each of their grandparents, and 12.5% from each of their great-grandparents. However, this is an average across many individuals, as each individual often receives more or less tested DNA markers from a particular ancestor than its littermate. Because of genetic factors, siblings can sometimes have differences in the breeds identified (or placement of the breeds) in their family trees due to differing actual percentages inherited from their grandparents and beyond. Essentially, each puppy received 50% of its DNA from their Dam and 50% from their Sire, however they will likely not inherit the same 50% as their siblings from either parent.  

The following website also has a short article discussing genetics of littermates: https://www.psychologytoday.com/blog/canine-corner/201108/why-are-some-litter-pups-uniform-in-appearance-while-others-are-mismatched

I have attached a copy of a statement that you are welcome to share with your customers, as well as the below links. Our goal is not only to provide the most accurate and robust information available to each of our customers, but also to make sure that this information is understood. If you have any further questions regarding how the testing works, or if your customers would like to discuss their results, we encourage you to contact our Customer Care Team.

https://help.wisdompanel.com/s/article/Wisdom-Panel-test-said-my-registered-purebred-dog-isn-t-a-purebred-How-can-this-be

https://help.wisdompanel.com/s/article/Do-the-Wisdom-Panel-tests-measure-the-purity-of-my-purebred-dog

https://help.wisdompanel.com/s/article/Can-Wisdom-Panel-certify-a-purebred-or-designer-dog

https://www.wisdompanel.com/how-it-works/breed-detection/

Regards,

Heidi

Wisdom Health