Scientific Studies on DNA Testing

Joined
Nov 17, 2012
Messages
540
Likes
1
Points
18
Location
Canada
#1
I volunteer with a rescue that does a fair bit of DNA testing - we rely on the tests to label some dogs as 'safe' or 'not safe' to be adopted within the province. So I find the topic of DNA testing really interesting.

A lot of people seem totally sold on the idea of DNA testing to determine what breed a dog is. I've had people show me their dog's DNA test and, although confused about the outcome, take it as fact. On the other hand, a lot of people find the tests totally laughable and claim they are a waste of money.

I'm on the fence. Leaning toward the latter group. I find it hard to wrap my head around it. Dogs are the most phenotypically diverse species on the planet. So... how... just HOW?

Are there any peer-reviewed, scientific studies to the accuracy of the tests? Or even any collected, organized anecdotal evidence of KNOWN mixes being tested?

So far, all of the information I have found to suggest the tests are actually fairly accurate has come from the labs or companies themselves, which is a red flag for me. Obviously they're going to be biased toward their own product and procedure.
 
Joined
Feb 26, 2013
Messages
290
Likes
0
Points
16
#2
I've had people show me their dog's DNA test and, although confused about the outcome, take it as fact. On the other hand, a lot of people find the tests totally laughable and claim they are a waste of money.
Just going to post this up here (not my dog) and you can work out which camp I'm in. :p



Are there any peer-reviewed, scientific studies to the accuracy of the tests? Or even any collected, organized anecdotal evidence of KNOWN mixes being tested?
Nope. Not that I'm aware of.
 

milos_mommy

Active Member
Joined
Oct 14, 2006
Messages
15,349
Likes
0
Points
36
#3
My understanding is this (although I'm not 100% sure where I heard it and if its true):

There are reliable DNA tests that can tell breeds, or at least very very specific types, in a dogs background. These tests are extremely costly (in the 1000$+ range) and are not widely available to the public, but they do exist.

The tests you can buy at the pet store or order online are highly inaccurate, and essentially a scam.
 

Saeleofu

Active Member
Joined
Oct 9, 2009
Messages
9,036
Likes
0
Points
36
#4
I think they CAN be accurate, but not all of them ARE. I did a test on Gavroche when I first got him(I forget the brand), and it said >90% Boxer with a hint of Scottish Terrier. I can see that, and accepted it as pretty darn accurate. I did another test on him last year since it was on sale for $1 (Wisdom Panel) and it had about 10 different breeds, only a bit of Boxer, and lots of things like Poodle, OES, etc that I can pretty much guarantee aren't there.

So...in short...it depends on the test. If it makes sense, go with it. If it doesn't, don't.
 

Airn

New Member
Joined
Nov 12, 2012
Messages
1,044
Likes
0
Points
0
Location
Bentonville, AR
#5
Aren't there a lot of breeds that aren't on the test?

I'm pretty certain that Gwen is a mix, but she is (mostly) Kelpie. So, if I take a test with Gwen and end up with some crazy results, couldn't it also be due to the fact that Kelpie is not an 'option'?

If it has any chance of being a realistic reading of what your dog's breed is, it would probably be a good idea to see if the breed you believe your dog mainly is, is even on their breed list.

But, I still haven't heard very good things about it. Plus, it's too expensive for shits and giggles, to me. :rofl1:
 

Laurelin

I'm All Ears
Joined
Nov 2, 2006
Messages
30,963
Likes
3
Points
0
Age
37
Location
Oklahoma
#7
Just look through these:

http://sheltermedicine.vetmed.ufl.e...udies/current-studies/dog-breeds/dna-results/

I don't believe the DNA tests at all. I mean, I know that dogs can come from surprising ancestry (ie: Squash and Shambles being half husky), BUT things like #33 where supposedly 4 solid colored dogs have given birth to an irish spotted dog.... that's impossible.

Or Dog #66 where supposedly 4 longhaired breeds produce a smooth coated dog?

No way would I be using DNA tests for live or die kinds of questions. Not at all.
 

MicksMom

Active Member
Joined
Sep 18, 2007
Messages
3,978
Likes
0
Points
36
Location
Warren Co, NJ
#8
Aren't there a lot of breeds that aren't on the test?

I'm pretty certain that Gwen is a mix, but she is (mostly) Kelpie. So, if I take a test with Gwen and end up with some crazy results, couldn't it also be due to the fact that Kelpie is not an 'option'?...
That. My favortie example of how inaccurate these tests are is when somebody from the old Acme forum had DNA testing done on one of her dogs a few years ago. It was very obvious by looking at the dog that she had a lot of Chow in her. The test results came back as the primary breed being Papillon. The only thing Pap about this dog was her coloring.
 
Joined
Nov 17, 2012
Messages
540
Likes
1
Points
18
Location
Canada
#9
So nobody knows of any scientific, peer reviewed studies then?

Thanks Laurelin and Frostfell for giving me some extra reading material.
 

Flyinsbt

New Member
Joined
Oct 6, 2012
Messages
886
Likes
0
Points
0
Location
Portland, OR
#11
Just look through these:

http://sheltermedicine.vetmed.ufl.e...udies/current-studies/dog-breeds/dna-results/

I don't believe the DNA tests at all. I mean, I know that dogs can come from surprising ancestry (ie: Squash and Shambles being half husky), BUT things like #33 where supposedly 4 solid colored dogs have given birth to an irish spotted dog.... that's impossible.

Or Dog #66 where supposedly 4 longhaired breeds produce a smooth coated dog?

No way would I be using DNA tests for live or die kinds of questions. Not at all.
I remember this thing, and it irritated the crap out of me. They used it as proof that people can't visually identify breed mixes, by using the unproven DNA test.
 

Laurelin

I'm All Ears
Joined
Nov 2, 2006
Messages
30,963
Likes
3
Points
0
Age
37
Location
Oklahoma
#12
I remember this thing, and it irritated the crap out of me. They used it as proof that people can't visually identify breed mixes, by using the unproven DNA test.
Annoyed the crap out of me too. Obviously some dogs hav surprising heritage based on their looks. However saying 4 long haired breeds birthed a short haired pup or 4 solid colored breeds birthed a piebald is... Yeah.
 
Joined
May 29, 2011
Messages
738
Likes
0
Points
0
Location
Kalamazoo.
#13
Magpie won the Wisdom Panel II in a raffle, otherwise I would never pay money for them. I always feel bad bursting bubbles when people get DNA tests and take to heart what their dog "is" based on the results.

Magpie's particularly was horse ****. She came back as an American Foxhound x English Spring Spaniel. To make it even funnier, they also say she has Puli, Beauceron, Canaan dog, English Cocker, and Doberman in her ancestry. Sure, OK.
 

milos_mommy

Active Member
Joined
Oct 14, 2006
Messages
15,349
Likes
0
Points
36
#14
There was a dog on here that weighed like 90 lbs, and his DNA test came back maltese X boston terrier.
 

Dagwall

New Member
Joined
Aug 28, 2013
Messages
277
Likes
0
Points
0
Location
Northern VA
#15
One of my coworkers did the Wisdom panel on both her dogs who are mixes. One looks like a chocolate lab mix and the other looks like a boxer/shepherd/lab type mix. I can't even remember what they came back as but neither really made sense for fit what they LOOK like. She now goes by the Wisdom panel results and not what they look like... She has a biology degree and works in a biotech lab with me... People will believe a lot even when they SHOULD know better.

I'd still like to do it myself just for sh*ts and giggles if I see a good sale on the kits but I certainly wouldn't instantly take the results as unquestioned truth.

 

Members online

No members online now.
Top