On March 4, CBC News (Canada’s publicly owned news and information service) ran a television and online report about an investigation of four companies that offer dog DNA tests that purport to identify the breeds in mixed-breed dogs. Within a day, practically everyone I know was posting links to the online report with comments like, “I knew those tests were bunk!”
In the past 15-plus years that they’ve been available, I’ve been skeptical of the ability of these commercial testing products myself. However, I will say that, in my experience – and that of the CBC report – two companies in particular seem to provide results that are at least in the ballpark of possibility for the most common dog breeds found in North America. And one company seems to have a pretty good handle on identifying the origin of mixed breed dogs from other parts of the world. My response to the report, though, takes in a few details that many commenters seemed to miss.
The CBC sent DNA samples for four individuals to four different companies that offer mixed-breed dog DNA tests: Accu-Metrics, DNA My Dog, Embark, and Wisdom Panel. But they picked odd (in my opinion) candidates to use for the tests: A human, a purebred Great Dane, a mixed-breed dog from Turkey, and a mixed-breed dog from Kuwait.
As far as the human sample was concerned: I was pleased to learn that Embark and Wisdom Panel immediately sussed out that no dog DNA was present in the sample. And was not terrifically surprised when Accu-Metrics and DNA My Dog returned various dog-breed mixes in their results for the human sample. (Before seeing this report, I had never heard of Accu-Metrics before, and, a long time ago, received similarly incredible results of a test from DNA My Dog.)
The latter two companies also failed to accurately identify the purebred Great Dane. Results from DNA MY Dog suggested the dog was mostly Great Dane, but also 10% -25% Staffordshire Bull Terrier. Accu-Metrics returned the breed that the CBC suggested on its submission form that the dog most resembled: a Chihuahua! I don’t see any need to further discuss any results – or ever recommend the services from – either of those two companies.
Both Embark and Wisdom Panel correctly identified the Great Dane as 100% Great Dane.
Mixed-Breed Dogs from Other Continents
I so wish that CBC had used mixed-breed dogs from North America as their last two “test dogs,” because there are likely to be very few representatives of the most common purebreds dogs on other continents in Embark’s and Wisdom Panel’s databases. The most common (or likely) mixed-breed dogs on the streets in Turkey and Kuwait are not likely to be the breeds that are most common (or likely candidates) mixed-breed dogs found in Canada or the U.S.
Wisdom Panel identified the breed mix for the Turkish dog as Segugio Italiano, Chihuahua, Anatolian Shepherd, German Shepherd, and Estrela Mountain Dog. Without information as to how common those dog breeds are found in Turkey, it’s impossible to know how accurate this might be. To its credit, Embark identified the breed mix of the same dog as 100% West Asian Village Dog – meaning they were able to pinpoint the mixed-breed dog’s geographical origins. I’d call that a home run!
Similarly, Embark identified the dog from Kuwait as 100% Arabian Village Dog – again, at least accurately identifying the dog’s geographical place of origin. (Kuwait is also located in West Asia, but also at the northern edge of Eastern Arabia.) Wisdom Panel identified the dog as being a mix of American Pit Bull Terrier, Chihuahua, German Shepherd, Segugio Italiano, and Xoloitzcuintle.
Comparing Embark and Wisdom Panel, the two leaders
I’m a little dubious about the idea that these two foreign-born dogs could share three breeds in their Wisdom Panel results (Chihuahua, German Shepherd, Segugio Italiano), so I’m rather more impressed with Embark’s performance here. However, I’d want to compare the results from these two companies on more prosaic mixed breed dogs from this continent before dismissing Wisdom Panel altogether. In our past comparisons, using my two mixed-breed dogs Otto and Woody, the results were pretty darn close.
Otto’s DNA Test Results:
Woody’s DNA Test Results:
Back when I was still fostering Boone, my 1-year-old adolescent dog, I ordered a DNA test kit from Wisdom Panel, and these were the results:
But I think I am going to go ahead and order a test kit from Embark, to compare these results. I have a feeling, based on the CBC report, that I might invest a little more confidence in Embark’s results, but I’ll let you know!
Based on the behavior and physical attributes of both our dogs, I feel Embark was pretty accurate. Especially with our Lab mix. He showed to be Lab/Chow Chow/Pit primarily. He acts very Lab like, loves water, carries things in his mouth. He is kind of built like a Lab too. Longer legs, thick tail. Weighs about 44 lbs. He can be very cautious, standoffish and leery of strangers, as I understand it, this seems to be kind of Chow like. Has kind of a curled up tail. And he sounds like a pitbull type dog. He has a certain way of vocalizing that sounds like other pits I have heard.
Our female, who is super anxiety ridden, sounds like a little “yippy” when excited, is a stomach on 4 legs, loves to track (her own version), dig and play in the snow. Her primary breed (50%) is beagle. Spot on with many things she does. Second (27%) is Border Collie. Um, yup. She yips and likes to “herd” someone she is playing with. I have had two vets (one specializing in behavior) say that is could be very likely that comes from the BC genes. Along with the anxiety. She can be suspicious of strangers and used to nip at peoples ankles before we worked on that behavior. She really looks like a combo of those breeds. (Super cute) She weighs about 33 lbs. She has a few other small percentages of other breeds, but their behavior does not stand out in her, at least to me.
So, I have to say Embark was pretty accurate. Knowing these facts did help us work on some of the less desirable things they did. These are two awesome dogs and knowing their primary breeds, in my opinon, has helped us manage them much better.
Funny, when we got our male, we were convinced he was a pointer mix, he was 8 mos., but he did not act like a pointer. A huge AHA moment when I got the DNA test results. For the money, I would say it was very much worth it for us.
Wish I had some money to throw into a Wisdom Panel test for each. Would be fun to compare the results between the two companies.
I did the DNA analysis of our poodle mix girl using Wisdom. Her results came back as “poodle mix”. 😀 One parent was poodle and the other parent was poodle and a rare type of poodle. I’ve never heard of a “rare” type of poodle, but I’m also not a dog breed expert. She looks like a 9 lb poodle and is a great dog. (Adopted at age 6 and she just turned 14 – and her previous owners were awesome but had to go into a nursing home. They have to be awesome because our girl is so well-mannered, etc.)
The good news for most of us is that the actual breeds that make our pups great don’t really matter, right? Although I can certainly see how these results might help rescues.
I did a lot of research before using Embark and then decided because they had some sort of link to Cornell College of Veterinary Medicine that I would use Embark. I have had four of my dogs’ DNA run by them. Two mixed breeds were fairly surprising. Two were pure bred, a Boxer and a toy Rat Terrier, both rescues. I did not send a picture of either and did not identify either as a pure bred. I had thought the toy Ratty was probably a chihuahua mix because of head shape and body shape. I have read that chis were sometimes mixed with standard Rattys to get the toy size but perhaps that is far enough back that it doesn’t show up in current testing. I liked the health reporting as well. I will likely continue to use Embark going forward.
I had Embark done on both my dogs. Wisdom Panel was done on one of them by the rescue before I adopted her. Embark says she’s a purebred, but WP said she was a mutt (5 different breeds & a 6th showing as ‘Mixed Breed’). She looks like a purebred, but who knows.
I am very happy with Embark not because of the breed results, but because my second dog came back as an AmStaff (or maybe it was AmPitt, can’t recall), but they were just starting to type the American Bully. After a few years, I started seeing frequent “close relatives” pop up for him that were all American Bully as their database expanded so I asked them to re-run his since they had a better genetic profile for that breed now & sure enough, they updated him to American Bully. It’s rare for any company to put in more work several years after the fact to correct results due to new information.
Also, we ran an age test on both dogs & embark notified us several months later that they have found inconsistencies in their age tests that they aren’t comfortable with so they refunded our money. I consider them a very reputable company.
7 years ago I used the
“DNA my dog test” on my fluffy white non-shedding stray who somewhat appeared like a Havanese mix. My husband suspected their validity when a photo of our dog was requested with the sample and the question: what do you think he is? Imagine our surprise when their results were “predominantly Pug”! When I complained to them this was wrong they apologized and sent a new test kit. The new test showed Australian Shepherd as predominant in the mix. Very doubtful still but I gave up on pursuing his identity and love him for whatever he is. That was a scam operation!
I feel Embark nailed my rescue dog’s composition. I sent them no photo, no info. Nothing except a swab. I was thrilled how accurate they were.
Also, the shelter est.his age to be 7 yrs. Embark did not know that. They said his genetic age was 8 yrs old. I love this company. I fully recommend them
DNA My Dog puts out their results in a somewhat confusing way for me with the range of percents. That said, I did my older dogs tests through them purely due to cost, and what they said makes sense for the boys.
One came back as up to 60% Labrador (and he DOES look incredibly Labby), then the other “half” was divided between Rottweiler (up to 25%), Bulldog (up to 25% ea), and Pit Bull (up to 9%). Now, I live in St. Louis, MO where certain breeds are more common than they are in other areas. Pit Bulls and bulldogs are more common than pure Labs. Rottweilers not so much, but I also don’t spend a lot of time in the backyard breeder community so I certainly can’t rule it out like the Great Dane in the CBC test group that was reported to the lab company as a chihuahua mix. He acts a lot more protective and is less orally fixated than the MANY Labs and mixes thereof I’ve trained through the years. He’s also got the Rottie curled tail up over his back, so I could see it. If he’d not been mistreated before we got him, it would be easier to resample him for another test, but I don’t know if we’ll redo his or not. Husbandry behaviors stress him badly.
My other dog tested through DNA My Dog came back as up to 60% Akita and the other half “up to 25%” Am. Staffordshire Bull Terrier, Bull Terrier, and Brittany. You’d expect a weird looking dog with that kind of result, and he’s a bit odd-looking, but he definitely does NOT look like an Akita. And yet, he’s a back-yard bred dog from St. Louis, and there are almost as many Akitas in the collective gene pool here as there are pits. So we have potentially more than 100% from both dogs, which means there’s some fuzzy math going on in more ways than 1. They may not be wrong. They’re just… odd.
I did Embark on 2 of my rescues. One came back 100% Malinois which I would believe. The other, I am saying Border Collie mix, came back as American Eskimo, mini Aussie plus other things which I did not agree with. My friend recently adopted a dog and did both Embark and Wisdom. Wisdom 65% Malinois 32% GSD and 3% Dutch Shepherd, Embark gave it 53% Dutch 27% GSD and 19% Malinois. So same breeds yet wildly different percentages
I did Embark on a rescue I got from a Miniature Schnauzer rescue. She has the color of a Schnauzer and the hair, according to my groomer (a Schnauzer breeder), is consistent with a Schnauzer mix. Her results returned as 45% Chihuahua, Shih Tzu, and some other breed. Miniature Schnauzer was not mentioned at all. I contacted the company to see if maybe it could be run again as her looks and behavior do not match that much Chihuahua. I was told, “We do not make mistakes.” That was the end of our conversation. Not good service and I’m not sure I trust the results.
I’ve done both on Diana pawPrints and Freyja grey. Aside from results, I like that Embark allows me to message relatives, something Wisdom panel does not. Diana has a brother, Pete, on Embark and they met for a play date. She has another brother on Wisdom, Winston, who we have no way of contacting.
As for Freyja Grey, I think her results on Wisdom are closer than Embark. I suspect she might be part Seppala Sled Dog and her results look very much like those of the breeds they’ve crossed into the working sled dogs.
Both Embark and Wisdom are working to improve their accuracy and I think Wisdom has come quite a ways as can be seen comparing Boone’s recent test to those of Otto and Woody.
Back in the infancy of these tests I had one done on Ramses. While I didn’t see it at first, once it came back and I saw his biggest parts were Afghan Hound and Boston Terrier, I started seeing those traits. The nose, his gait when he ran towards me.
I used both Wisdom Panel and Embark. Wisdom Panel reported that my black and tan 55# dog was chihuahua and dalmation. My dogs black and tan coloring is similar to a doberman or rottweiler. Then I tried Embark. Embark came back with beagle, german shepherd, lab, retreiver, chow, aussie, collie. His personality is very much in line with the beagle. He likes to run off and explore if he gets a chance and has done very well in nosework. I have used Embark with another rescued dog and a friend used Embark on a rescued puppy mill breeder. My opinion is Wisdom Panel missed the mark completely and Embark has been excellent in all respects. I would recommend Embark and only Embark.
The shelter thought Duncan was a Papillon mix. Probably bc of his butterfly fringed ears. He is 40% long haired chihuahua. Fifteen % Pomeranian. Hence the ears. There are so many wonderful things about Embark that I could go on and on but I won’t. I loved the fact I sent them only a swab. No pic, absolutely no info at all. Just a swab and they nailed it. The results explained so many questions I had about him. I would never use anyone else. They send me other dog’s info that have the same mix and give me the opportunity to touch base with those people. I take an annual survey to help in their research
My rescue looks like a Rat Terrier, but at first they said Jack Russell Terrier. So I did Wisdom to see and it came back 50% Parson Russell Terrier and 50% companion dog. That was almost helpful! Not that it really matters to me, as I love him to pieces.