Wait, my dog’s DNA says he’s what? Can that be right?
October 16th, 2009
|
| Share |
This is my dog, Tucker. Tucker is a male dog that I adopted from a shelter near me. His history is unknown, as he was found as a stray. When I got him he was estimated to be between six months and a year old, which would make him about two and a half or three today. The vet was confident he was not signifficantly older than a year old. However, we had no idea of the breed mix.

Tucker is very small, about ten pounds. His face looks distinctly terrier like and his habits are also very terrier-like. He’s prone to jumping, he likes to hide things like bones and come back for them later and he’s very dominant toward other dogs, even larger ones, although he is not generally aggressive.
I figured him to most probably be a combination of a Chihuahua and a Border Terrier. He looks more like a border terrier than any other kind of terrier, Border Terriers are not very much larger than he is and they seemed to be the closest match. They have wire hair, like he does and many of their disposition traits match him. However, his ears look very chihuahua-like and he has big eyes like a Chihuahua. He is also quite small, a bit smaller than most terriers, but he’s about the right size to be a mix of the two.
Some Chihuahuas are of the “hairless” variety, which have very little fur. Others have short hair which is of a distribution and length similar to that of Tucker, although the texture is different. The color is consistent with either breed.
Others generally had a similar opinion. The veterinarians that have seen him universally agreed he was a terrier mix of some kind. They also said that he could quite possibly be a Chihuahua-Border Terrier mix. The general agreement was that he likely had some Chihuahua in him. Some of the vets and veterinary assistants suggested that he include a different breed of terrier than Border, citing the fact that they’re al that common of a breed.
So there we have it: He is primarily terrier, possibly Border or another small breed terrier and also shows traits similar to a Chihuahua or possibly a mix of other small dog breeds. Right? Wrong.
Out of curiosity and because it’s only about fifty dollars, I sent out his DNA to be tested for breed. It’s a relatively simple affair. You get two cotton-tipped swab sticks that you rub on the dogs gums and inner cheeks and place in envelopes to send off for testing. Boy did I get a shock when the results came back!


Okay so he’s… wait… what the…?
A “Level 2″ result is a fairly strong match. It could indicate that one of the dog’s parents are of the type noted. It would also be consistent with at least one of the grandparents being a purebred. One would expect with that level of a match that either one or more of the grandparents were nearly 100% of the breed in question or that multiple grandparents had at significant portion of the breed’s heredity.
The Newfoundland is a very large dog, with normal adult males almost always being over 100 lbs and often 150 lbs. They’re sometimes called “gentle giants” for their mild temperament and extremely large size. They have very broad faces and large heads. The hair of the Newfoundland is usually black and they have a real lot of hair. The hair is generally very thick and has a substantial undercoat. They’re one of the strongest and largest breeds of domestic dog.
Could it be that this is a mistake due to some other breed of dog being closely related to the Newfoundland? Unlikely. The Newfoundland breed has been around for a long time and was bred from the dogs native to the area of Newfoundland. Their closest relatives amongst domestic dogs are Laboradore Retrievers, but labs are common and therefore would likely have had their genetics well indexed.
As for the other breeds represented, a level three is still fairly high. It could indicate that a grandparent was a purebred. It is likely that at least one of the grandparents had a strong inheritance of the given breed or that more than one had some level of inheritance from the breed. A level four could indicate a grandparent that is a purebred, but is more likely a great grandparent.
The other dog breeds identified include:
Maltese – That one seems to make some sense. The Maltese is a small companion dog that is about the right size. It also does have some features that Tucker has, including large eyes and some facial features. They only weigh four to ten pounds. The hair is completely wrong for Tucker, but given that it’s only a level three, it could easily be explained as simply being a trait that was not inherited from that ancestor.
Papillon – The Papillon is a small dog of the Spanial type. While I would have never guessed this to be one of Tucker’s ancestors, it does make sense. The Papillon is fairly small in size, and has ears that are similar to the ones that he has. The eyes and face do show some resemblance and the dog is known for “thinking it’s a big dog” – meaning that it tends to be bold and dominant. It is also fairly athletic and has good jumping ability.
The Palpion is also a fairly well established breed of dog that is unlikely to be mistaken in a genetic test. However, if there were another breed, it would likely be a Spaniel. No other members of that family appear to be a better match.
Mastiff - Mastiff is sometimes used to describe a group of breeds also known as Molossers. However, the proper Mastiff is a breed classically represented by the English Mastiff, but also includes the American Mastiff and the Spanish Mastiff. These are amoungst the largest domestic dogs around. Although Great Danes are generally taller, adult English and American Mastiffs can be up to 250 pounds.
Mastiffs have been used as sheep dogs, cattle dogs and guard dogs, due to their very large size and robust nature. They’re not a very popular breed, at least not in recent years, although there are a reasonable number of them.
It’s highly unlikely that a much smaller dog could be mistaken for a Mastiff on a DNA test. Their closest relative that is of a small size is the Pug, but Biopet labs has the Pug already in their database.
Chinese Shar-Pei – Another real shocker on this one. There are few dogs I can think of that look LESS like Tucker. The Shar Pei is a medium to large sized dog that is believed to be related to the Chow-Chow. The average adult weighs about 50 lbs. The breed is known for often having a distinctive “wrinkled” look, although not all individuals display this to the extent of the one pictured.
I’ve got no idea about this one either. I never would have guessed this.
Given, it is only a level four match, but it still seems to have almost nothing in common with Tucker. It’s possible that the short hair trait came from this breed or that none of the distinctive traits are readily apparent.
So just what the hell is going on here? Two of the dogs seem to be a reasonable possibility but the other three are about as different as I could imagine. Could it be contamination? It certainly was not on my side, as I was both very careful with the sample and would have no way of getting material from the dogs mentioned above onto the sample. I’d have to assume the lab has procedures to isolate samples and avoid contamination.
If this is real, that brings up another issue. Just how do dog breeds of such vastly different size… get lucky. This is actually more than an issue of… mechanics and geometry. A small dog could father dogs with a larger dog, but if the mother is much smaller, then the offspring will almost certainly be born too premature to be viable. A small dog is simply not capable of providing enough space in the uterus for the gestation of a large breed dog.
Of course, size is not controlled by a single gene and it can be a little chaotic. A small and a large dog don’t always produce a medium sized dog, although in most circumstances the offspring is somewhere between the sizes of the two parents.
It’s hard for me to even conceive of a hypothetical situation that could cause this, short of human intervention and artificial insemination. Could an abnormally small Maltese Shar-Pei mate with an abnormally large Maltes? Or a large Shar Pei mated with a Mastif, resulting in a dog that was roughly the size of a Shar Pei, which mated with a large Papillon? The presence of the Newfoundland material in such high levels is the hardest part to reconcile.
If anyone knows anything about dog breeding or genetic testing and thinks they can offer some insight, I’d love to hear it!
Since the test was reasonably cheap, I’m considering doing it again, just to see if it might have been contamination.
This entry was posted on Friday, October 16th, 2009 at 7:02 pm and is filed under Good Science, Misc. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.
View blog reactions




October 16th, 2009 at 10:06 pm
LOL – that Maltese made me laugh out loud. Looks like a mop with a face.
Quote Comment
October 16th, 2009 at 11:32 pm
I would be VERY skeptical of the results. There are a lot of careful procedures that need to be followed to assure that samples are isolated and not contaminated. Whoever opens the envelopes should do it in an isolated area and then divide the sample up and clean the area and change their gloves and any tools between each sample. They should divide the sample and have two or more portions handled separately to confirm the results and the test should be done alongside control samples to assure the accuracy of the results.
I very much doubt that they could do all this and still make a profit on a test that cheap. To not lose money on such a cheap test I doubt they could really implement all the labor-intensive procedures that you’d find in a forensic or research DNA laboratory.
It is VERY easy to cross contaminate DNA when the samples are small and many are being handled.
Quote Comment
October 16th, 2009 at 11:54 pm
I read somewhere about 10 years ago or so, sometime around when genetic testing started to become common, that somebody had the bright idea to test various dog breeds to see if they could tell how old they were, by the amount of genetic divergence. However, when they started collecting samples, they discovered that there was more variation between dogs of the same breed than there was between different breeds.
On that basis, and that the results you got seem to be nonsense, I’d say they’re basically a crock. That’s not to say that the lab didn’t actually compare the sample you sent to samples from the breeds they mentioned, just that those comparisons don’t actually mean anything. Or maybe they knew that and realized nobody could argue if they just pulled names out of a hat.
I do think it would be interesting to send in another sample, preferably with different names, and see what the results are, but I don’t see it actually telling you anything about your dog’s ancestry.
Quote Comment
October 17th, 2009 at 2:19 am
Uh.. that dog does not have newfoundland in him, or if he does it has to be several generations back. It does not have a close relation to Newfoundlands and Mastiffs.
Those are two of the top ten largest breeds of dog and that guy is tiny. It could not be more off than that.
I don’t know what went wrong, but something sure did.
Quote Comment
October 17th, 2009 at 2:29 am
I’m afraid based on what little I know of genetic analysis, I have to agree with grmann. This looks like the pet equivalent to those services that purport to do a full genealogical history of your family for $500.
Quote Comment
October 17th, 2009 at 7:51 pm
No one brought up another obvious form of human error.. what if your sample got switched accidentally and someone else has a huge dog whose certificate says terrier/chihuahua now.
Quote Comment
October 17th, 2009 at 8:05 pm
Born again Skeptic said:
I thought of that and that’s possible, but still, I think there are issues with the fact that it would be a bit weird to have that mix in any dog due to the size differences.
I really have just concluded the test is probably bunk.
grmann said:
That doesn’t surprise me too much, but there would have to be some distinct differences that could be detected. Many breeds have distinctive traits, so certain genes would have to be in common. Most of the genes in any organism do stuff like create cell membranes and enzymes necessary for respiration and such things. Therefore, if you look at the whole genome there might not be any apparent similarity between the members of the same breed. However, if you focus on the portions that would have common characteristics, then there should be things that would stand out.
I think the best way to do it would probably be to seperate each of the chromosomes in a sample, since each one is inherited independently and then trying to match each type to a breed by looking for any portions unique and consistent within the breed. There would certainly be some.
Then of course, there’s the mitochondrial DNA which could give the dog’s maternal heritage. You could tell how far it deviated from the established samples due to mutations since mitochondrial DNA otherwise stays the same and is inherited directly from the mother.
But you could never do this for such a price.
Quote Comment
October 18th, 2009 at 12:21 am
I;m sure you could find out about a dog’s heridity through the DNA, but it would require more extensive tests and I don’t know that we yet have enough data on the genetics of various dogs to even begin to look for the common genes (We may have enough or we may not, I just don’t know.)
I do agree though that it would be a bit more involved than a $50 test.
Quote Comment
October 19th, 2009 at 4:01 am
Joe said:
true, wonder if you could train it to mop floors, had been a very good idea, just heard about a course to train your dogs to search for mushrooms.
Quote Comment
October 20th, 2009 at 8:38 pm
Papillon means butterfly. Perhaps that’s where your dog got his small size. Do we have any compelling evidence that one of his ancestors wasn’t an insect?
Quote Comment
October 25th, 2009 at 2:56 am
Hi! I’m a veterinary technician and I do volunteer work for my local animal shelter. As you can imagine, I see a lot of dogs on a daily basis.
I’d have to agree that this “Breed Identification Test” is inaccurate. Your little dog definitely has some sort of terrier in him. It is quite apparent in his size, shape, and wire-haired textured coat.
The ear shape is tricky. I do see where the Papillon could come in here. The Maltese is a possibility, but I think even that is a slim chance. Possibly a parent could have been a Maltese Mix.
The rest of the breeds listed (especially at the percentages given) are definitely not accurate. The only way to find out how accurate the testing is would be to have someone send in a sample of a dog where the ancestry is already known.
Either way, Tucker is absolutely adorable!
Quote Comment
April 21st, 2010 at 2:52 pm
Another skeptic
I also sent in my dog’s DNA and got a strange result for my 60 pound blue-tongued chow cross
They are redoing for free – no results yet
Level 2 Chihuahua
Level 3 Pomeranian
Level 3 Bull Terrier
Level 4 Brittany Spaniel
Level 4 Golden Retriever
Quote Comment
April 30th, 2010 at 2:05 am
I did a test as well for my 3rd generation 70lb mutt. I was thinking about having them test her father to see the accuracy and reliability of the test. I’ve always told everyone that she was a Rottweiler/German Shepherd/Chow Chow mix. However, only one of those breeds is in her according to BioPet. Here’s her photo:
http://s267.photobucket.com/albums/ii301/brittanyv326/?action=view¤t=100_1870.jpg&newest=1
Her results?
Level 2 Labrador Retriever (I could see this… as like a base coat of paint kind of thing…)
Level 3 Chow Chow (Least I was right about something – her tail curves and she has an undercoat)
Level 4 Chihuahua (What? This is 10-19% range… seriously?)
Level 4 Dachshund (Again, WHAT? Average weight is like 8-11lbs for these little guys)
Level 4 Dalmatian (Is this where she gets the slim frame as compared to Chows and Labs? I don’t know)
Quote Comment
May 22nd, 2010 at 6:26 pm
hellooo!, thanks 4 the information, this post was really nice ! I just received a copy of a good Mastiff Book. Wow! I recommend it for anyone thinking of getting a mastiff !
Quote Comment