Interesting Survey Results on Dog Breed ID

[Tco87]

---

Members do not see this ad.

A while ago, several people at the shelter I worked at participated in a study held by UF's Shelter Medicine Program. It asked "experts" (shelter workers, vets, vet techs, etc.) to identify a dog's predominant breed based off of pictures and details about their size and age. They then compared the survey results to the results of DNA tests.

And there is a HUGE lack of correlation. The results are here: http://sheltermedicine.vetmed.ufl.e...udies/current-studies/dog-breeds/dna-results/

There were so many dogs where the breeds identified by the genetic testing didn't match any of the top survey responses. I found it really eye opening because at the shelter I work at we do have to guess what breed a lot of dogs are based off of their appearance. I wonder how accurate we've been......

Anyway, I just thought this was interesting and wanted to see what other people thought.

 

[Whyevernot55]

I guess my question is about the validity of the DNA test itself - one of my vets recently showed me an article from VIN about the various DNA tests out there for dogs (ie the Wisdom Panel) and how accurate it is. So without knowing what tests they used and what the accuracy rate of it is, I have to take the survey results with a grain (or pound) of salt. My personal experience with the Wisdom Panel leads me to believe it's not as accurate as many people think, and I tend to be skeptical of the results.

 

[Bracco Pointer]

I guess my question is about the validity of the DNA test itself - one of my vets recently showed me an article from VIN about the various DNA tests out there for dogs (ie the Wisdom Panel) and how accurate it is. So without knowing what tests they used and what the accuracy rate of it is, I have to take the survey results with a grain (or pound) of salt. My personal experience with the Wisdom Panel leads me to believe it's not as accurate as many people think, and I tend to be skeptical of the results.

I agree. Especially with some of the breeds listed in large percentages. Ie: the Harrier. This is a rare breed dog that is bred only minimally in the US. The likelihood of one of these dogs running loose and breeding strays is low. Very low.

The study is still fun to look at though! 😀

 

[OwlandPussycat]

I guess my question is about the validity of the DNA test itself - one of my vets recently showed me an article from VIN about the various DNA tests out there for dogs (ie the Wisdom Panel) and how accurate it is. So without knowing what tests they used and what the accuracy rate of it is, I have to take the survey results with a grain (or pound) of salt. My personal experience with the Wisdom Panel leads me to believe it's not as accurate as many people think, and I tend to be skeptical of the results.

I've also heard some not-so-stellar things about the DNA tests. When people ask about them at my clinic, the vets tell them it's not worth it.
I'm curious as to whether control-type studies have been done. Like, has anyone taken mixed breeds with known ancestries and submitted their DNA?
Anyways, in a shelter setting, I think that the biggest benefit to ID'ing the breeds in mixed dogs is to help potential owners predict things like behavior and adult size. Visual ID'ing by staff might be more useful in those respects.

Don't let this study make you doubt yourself!


Edit: so looking down the list more closely, it kind of seem like the more, er, "mixed" a dog is, the weirder the results are. For example, I don't think there's any way that Dog #25 is part bernese, rhodesian, and entlebucher sennenhund. On the other hand, american bulldog/american staffordshire for Dog 9 looks like its right on the money. I also see a lot of rare breeds listed, which seems kind of unlikely to me. But hey, you never know, right?

 

[krist]

I totally agree with the weird results of the "DNA test" like you guys have said, I highly doubt there are a bunch of Dandie Dinmont Terriers (WTF that's a breed?), Sealyham Terriers, or German Spitz's running around loose makin babies. Also, holy ears on dog 13!

Guessing mixed breeds is always difficult, especially on puppies, although it helps if sometimes they come in with the mom or if the owners know that dad was probably the neighbors intact male.

 

[wildcatj]

I totally agree with the weird results of the "DNA test" like you guys have said, I highly doubt there are a bunch of Dandie Dinmont Terriers (WTF that's a breed?), Sealyham Terriers, or German Spitz's running around loose makin babies. Also, holy ears on dog 13!

Guessing mixed breeds is always difficult, especially on puppies, although it helps if sometimes they come in with the mom or if the owners know that dad was probably the neighbors intact male.

I can't get over them. WANT!!!😍😍

 

[PetPony]

Owland, I read an article a little while ago where they tried the DNA tests on mixed breeds where they knew the parents. It was very small though, so not many dogs were tested. It showed that the DNA test weren't very accurate or even always gave the same results on the same dog.

I like how #106 apparently has Husky in it..

 

[krist]

Another weird thing about this survey is where are all the pitbulls?? That's one problem with a lot of the DNA tests I've read/heard about is there is no DNA available for a pitbull (because there's hardly any standardization of the breed at all) but where I live almost all of the strays and litters have at least some pit in them...and I find it weird that none of those dogs listed had pitbull even in the survey results? And the dogs that looked really pitt-y to me had Am Staff Terrier or something like that but that is not the same as a pitbull.

 

[PetPony]

The DNA tests don't have pit bull DNA available because of what you said (that's what I heard at least), so maybe they keep the pit bull out of the guessed results, too? 😕

 

[orcagirl]

I just want to say when people look at a dog just because it is black and white (tuxedo-style), it does not mean the dog has Boston terrier in it.

 

[orcagirl]

Another weird thing about this survey is where are all the pitbulls?? That's one problem with a lot of the DNA tests I've read/heard about is there is no DNA available for a pitbull (because there's hardly any standardization of the breed at all) but where I live almost all of the strays and litters have at least some pit in them...and I find it weird that none of those dogs listed had pitbull even in the survey results? And the dogs that looked really pitt-y to me had Am Staff Terrier or something like that but that is not the same as a pitbull.

I've always heard that Am Staffs are like smaller pit-types. Granted this comes from an area where 50% of the canine people probably has some pit in it.

 

[PetPony]

So what is the difference between them? I tried googling it but everybody says something different. What I heard the most is that ASTs have certain requirements to be met (for AKC etc) and PBs do not.

 

[orcagirl]

So what is the difference between them? I tried googling it but everybody says something different. What I heard the most is that ASTs have certain requirements to be met (for AKC etc) and PBs do not.

From google, I'm getting that pit bull is a general term to describe a type of dog, while ASTs are just one breed that falls under the pit term.

 

[primadonna22274]

Adorable. Any guesses on what my little guy Henry is? I debate having him tested but have also been unimpressed with the reliability of DNA tests. He was a stray, is about 9 yr old (was about 7 mos by dental exam when I adopted him). Excellent athlete. Not too smart. Very loving and only a bit jealous when his sister tries to take his favorite toy. Does NOT like water.
View attachment ImageUploadedByTapatalk1342969872.759223.jpgView attachment ImageUploadedByTapatalk1342969917.511856.jpg

 

[krist]

Yeah sort of, am staff is an actual akc breed and a pitbull is supposed to be an actual breed, but is so overbred and mixed with other things that it's become an umbrella term to mean anything that looks like that. Very few people know what a pitbull is supposed to look like.


Sent from my iPhone using SDN Mobile app

 

[kakurubird]

This article was posted on here somewhere a while ago and they test some mutts, a pure-bred dog, and some mixed breeds of known parentage.

 

[PetPony]

This article was posted on here somewhere a while ago and they test some mutts, a pure-bred dog, and some mixed breeds of known parentage.

That't the one I referred to earlier. Thanks for finding it.

 

[devyn]

I've always heard that Am Staffs are like smaller pit-types. Granted this comes from an area where 50% of the canine people probably has some pit in it.

This is wrong.

The American Staffordshire Terrier and the American Pit Bull Terrier were originally the same breed, but in the 1930s the breed split off into two separate breeds recognized by two separate kennel clubs. The AmStaff by the AKC and the APBT by the UKC. The AmStaff became a dog bred for conformation and temperament - they were bred for the show ring predominantly. The APBT was bred predominantly as a working breed. The dogs are generally around the same size, but there standards vary a little bit for weight and color allowances.

The Staffordshire Bull Terrier is a much smaller dog and is more popular in the UK, Ireland, and Australia. It is not as easy to find them in the states. They top out at around 34 lbs full grown for the males and 16" tops height wise. They are sturdy, small dogs and very good with children - this is the dog known commonly as the "nanny dog". Dogs in shelters are commonly mislabeled as SBTs.

Here is a link with more info - for anyone interested -
http://www.pbrc.net/faq.html

Just wanted to jump in here and clear that up - there are a lot of misconceptions about the different breeds under the term "pit bull".

 

[orcagirl]

This is wrong.

The American Staffordshire Terrier and the American Pit Bull Terrier were originally the same breed, but in the 1930s the breed split off into two separate breeds recognized by two separate kennel clubs. The AmStaff by the AKC and the APBT by the UKC. The AmStaff became a dog bred for conformation and temperament - they were bred for the show ring predominantly. The APBT was bred predominantly as a working breed. The dogs are generally around the same size, but there standards vary a little bit for weight and color allowances.

The Staffordshire Bull Terrier is a much smaller dog and is more popular in the UK, Ireland, and Australia. It is not as easy to find them in the states. They top out at around 34 lbs full grown for the males and 16" tops height wise. They are sturdy, small dogs and very good with children - this is the dog known commonly as the "nanny dog". Dogs in shelters are commonly mislabeled as SBTs.

Here is a link with more info - for anyone interested -
http://www.pbrc.net/faq.html

Just wanted to jump in here and clear that up - there are a lot of misconceptions about the different breeds under the term "pit bull".

Thanks. 🙂

 

[Whyevernot55]

This article was posted on here somewhere a while ago and they test some mutts, a pure-bred dog, and some mixed breeds of known parentage.

Yup, that's the article I saw from my vet.

Personal experience says the DNA tests can be fun, but can be a total waste. My old boss's dog, who basically looks like a Jack Russell but is 35lbs, came back as primarily Rott. Our office manager's lanky, tall, 60lb mutt (definitely some hound in there!) came back with lots more Chihuahua than should be possible. We had some come back that make total sense, but that was less common. The deal breaker for me was when someone called and complained that their "purebred Aussiedoodle"'s test came back as something other than Aussie and Poodle, and the company changed the results. 🙄

 

[Trilt]

So in short, we're all like:

...right?

 

[storyofmylife]

Even though though the results on many of the dogs are a little odd to me, it was so cool to read through them! Thanks for posting! What fun research!

 

[Tco87]

So in short, we're all like:

...right?

Hahaha, pretty much!😀

Thank you Kakuru, for posting that article. The survey results make more sense now. I did think a lot of them seemed weird and unlikely, but since there were others that made sense to me I just believed that the tests were more accurate. Also, we're seen weird pairing of breeds at the shelter I work at.

That makes me wonder what the point of the survey was though. I mean, if the tests are as inaccurate as that article makes them seem, how do we know which results we can trust? I supposed the take home message is still the same, don't assume you can guess all the breeds in a dog's ancestry off of it's appearance.

 

[OwlandPussycat]

Hahaha, pretty much!😀

Thank you Kakuru, for posting that article. The survey results make more sense now. I did think a lot of them seemed weird and unlikely, but since there were others that made sense to me I just believed that the tests were more accurate. Also, we're seen weird pairing of breeds at the shelter I work at.

That makes me wonder what the point of the survey was though. I mean, if the tests are as inaccurate as that article makes them seem, how do we know which results we can trust? I supposed the take home message is still the same, don't assume you can guess all the breeds in a dog's ancestry off of it's appearance.

I guess the take-home message is that we have a long way to go in the development of the DNA tests. Until we get there, we'll just have tell our clients to save their money instead of testing their dogs.

 

[Tco87]

I guess the take-home message is that we have a long way to go in the development of the DNA tests. Until we get there, we'll just have tell our clients to save their money instead of testing their dogs.

That too.

 

[orca2011]

I guess the take-home message is that we have a long way to go in the development of the DNA tests. Until we get there, we'll just have tell our clients to save their money instead of testing their dogs.

Haha. I've told clients this several times already. I love when they come in "I found this dog. Can you tell me what breed it is?" Go to answer in my area is a chi x or pit x of some sort.

 

[Tiger369]

So in short, we're all like:

...right?

This pic is soooo cute! Anyway, I hope the DNA testing gets alot better in the future with a lot less mistakes.

 

[OneDay81]

I'm curious as to whether control-type studies have been done. Like, has anyone taken mixed breeds with known ancestries and submitted their DNA?

I'm pretty sure that's how they built the data base. For what its worth, wisdom panel tested my dog and it was right in the money ( half lab, half chessie).
The "newer" the breed, the weaker the DNA will be, so to speak, so the more chance of the results not being as accurate, especially if its a mix of weak breeds. Even labs were mutts at one point.

 

[OwlandPussycat]

I'm pretty sure that's how they built the data base. For what its worth, wisdom panel tested my dog and it was right in the money ( half lab, half chessie).
The "newer" the breed, the weaker the DNA will be, so to speak, so the more chance of the results not being as accurate, especially if its a mix of weak breeds. Even labs were mutts at one point.

I was noticing that in the shelter survey, the dogs who seemed to have fewer breeds mixed in also seemed to have results that were more reasonable. I wonder if the degree to which a dog is "mixed" makes a difference. It also sounds like the wisdom panel is the more reputable of the available ones. I would really like to see it fine-tuned further, because I think it's a really cool concept!

 

[Trematode]

I wonder what genetic changes constitute a new breed.

For example, I am currently working in a lab that wants to sequence the genome for every apple cultivar. Characteristics like colour and taste is caused by SNPs (single nucleotide polymorphisms). So a difference in one nucleotide in the entire genome can change the apple from red to green for example. If breed characteristics were caused by SNPs, they may be easy to find. But a lot of characteristics may be caused by changes in multiple genes. With apples, traits like disease resistant may be caused by changes in many genes. We are trying to figure out how accurately you can predict traits by looking at the genome.

So I wonder how they determine breeds.

I'm comparing apples to dogs though....

 

[GreenMouse]

The Wisdom Panel developers came through and did a talk with our shelter club this past year. I ended up running the test on my two dogs and the results were not too wildly unbelievable. I agree with the previous posts - the newer the breed the less accurate the results. They try hard to find unique markers for the breeds, but if two breeds use the same foundation, it becomes difficult to tell them apart. One funny thing the researcher that did the talk mentioned is that apparently the chihuahua and great dane have a lot of very similar markers, so they sometimes have trouble telling those two apart.

Here are the results from my girls:

Freya is about 55lbs, very active, and very smart

She came back as: 25% American Bulldog, 25% Miniature Bull Terrier, 12.5% Samoyed, and everything else too mixed to be able to say.

Ripley is about 45lbs, very active, not as smart as Freya but still learns quickly, loves people and is very willing to please.

She came back as: 25% American Bulldog, 25% American Staffordshire Terrier, 25% Kuvasz (this is the one I find weird), and everything else mixed.

I had a lot of fun doing the tests on them. I definitely take them with a grain of salt, but in my case they were not too crazy.

 

[PetPony]

Ripley is adorable! I really want a brindle dog whenever I decide to get one to keep permanently.

 

[GreenMouse]

Ripley is adorable! I really want a brindle dog whenever I decide to get one to keep permanently.

Awww - thanks! She is a cutie and knows how to lay on the charm. She is actually a foster and is up for adoption 😉
(I have no subtlety...)

 

[PetPony]

Awww - thanks! She is a cutie and knows how to lay on the charm. She is actually a foster and is up for adoption 😉
(I have no subtlety...)

I don't feel comfortable enough to have my own pets yet because I don't make enough money (I don't judge anyone else that does though. I just don't want my own). That's why I raise a guide dog (don't have to pay anything but food) and will be fostering next year. But later, once I'm comfortable to get my own, I want a brindle. But we'll see what dogs I get to meet at that time then and it might be one that looks completely different. You never know.

 

[GreenMouse]

I don't feel comfortable enough to have my own pets yet because I don't make enough money (I don't judge anyone else that does though. I just don't want my own). That's why I raise a guide dog (don't have to pay anything but food) and will be fostering next year. But later, once I'm comfortable to get my own, I want a brindle. But we'll see what dogs I get to meet at that time then and it might be one that looks completely different. You never know.

Good for you for knowing what you are able to commit to (and for fostering)! When my husband and I went to adopt we were planning on coming home with a lab or a golden, and ended up with Freya. Don't regret it for a second - but definitely not what we set out planning to find 🙂