Frequently Asked Questions

On May 21, 2003, investigators at the FHCRC (and now at the NHGRI) published a paper in the journal Science on the population genetics of purebred dogs. Science subscribers can find the paper in the May 21st 2004 print edition or online using the links provided on our homepage, http://research.nhgri.nih.gov/dog_genome/. For non-subscribers, most public libraries carry Science, so you can also find it there. The reference is: Parker H.G., Kim L.V., Sutter N.B., Carlson S., Lorentzen T.D., Malek T.B., Johnson G.J., DeFrance H.B., Ostrander E.A., and Kruglyak L. (2004) Genetic Structure of the Purebred Domestic Dog. Science, Vol 304, Issue 5674, 1160-1164 , 21 May 2004
Several people have written in with comments or questions about this paper. The most frequently asked questions and responses are as follows:

Can you tell me the most important findings from the paper?
There are three primary findings in the paper. The first is a demonstration that dog breeds, as predicted, are generally closed breeding populations and that any given dog is more closely related to members of its own breed than dogs of any other breed. Second, and as a result of the first, in a blinded analysis we were able to correctly assign 99% of individual dogs correctly to their breeds. Finally, genetic analysis clustered the 85 breeds into four main populations reflecting similarities in morphology and geographic origin.

Does that mean you can take a cheek swab and determine what breed a given dog is?
For the breeds in the study (85 AKC recognized breeds) we can do that with roughly 99% accuracy. Our accuracy will probably improve even beyond that as we add more dogs to the database. However, this is something we are not currently offering as a service. We would not be able to accurately determine what breed a cheek swab sample came from for breeds that were not included in our study.

Can you tell what the origins of mixed breed dogs are?
It's a great question and an obvious outgrowth of what we've done to date. We're hard at work on it now, but no data on mixed breed dogs is included in our Science paper.

The paper reports that four dogs failed to be assigned to their breed groups. What does that mean?
It does not mean a great deal in the overall scheme of the study. We found that four of 414 samples we analyzed, those from a single Beagle, a Chihuahua and two German Shorthaired Pointers were not assigned correctly to their breed in a blinded computational analysis. While the result may reflect something odd about those particular dogs, it's important to keep in mind that the study involved hundreds of DNA samples and we receive hundreds of samples a year from dog owners. Samples could have easily been mislabeled by owners or veterinarians, or mixed up by people doing the initial processing. Importantly, for the Beagle, Chihuahua, and German Shorthaired Pointer, other samples we analyzed from these breeds were all correctly assigned to their breed by the computer.

One of the four groups in the paper are of "Ancient Origins." What does that mean?
Many dog breeds can trace their origins back hundreds if not thousand of years. The breeds specifically assigned to this group showed similarity at the DNA level to grey wolf and each other, and hence are labeled as "ancient". This group included many of the Asian breeds (Akita, Shar-Pei, Shiba Inu), some hounds (Saluki and Afghan Hound), and some of the Spitz type dogs (Samoyed, Siberian Husky, Alaskan Malamute).

The Ancient group did not include the Pharaoh Hound, Ibizan Hound or some other breeds that we believe have been around for thousands of years. Why? Does that mean these breeds are so called "modern recreations"?
Not necessarily, and this is an important (and frequently misunderstood) point. In all likelihood for many breeds, if the records were available, one could track continuous lineages of thousands of years and yet they didn't end up in the "ancient group" in our paper. Those breeds may represent modern recreations of old breeds. But it is just as likely that admixture of other breeds in the intervening thousands of years has blurred the DNA fingerprint of those breeds to the point where currently available tools can't detect the ancient lineage. When did that admixture occur? Our study does not address that.

So does that mean that breeds like the Pharaoh Hound or Ibizan Hound could retain an unbroken line dating back several thousand years?
We truly don't know. Our study does not tell us. But contrary to what most of the media has claimed, our data is not at all inconsistent with that interpretation.

What does this study say about behaviors like aggression?
Nothing. The study does not attempt to localize any genes for any behaviors.

The Whatley Review published an online article saying our study found that Chihuahuas were not really dogs. Is that true?
No! Our study did no such thing. Of course Chihuahuas are dogs as are all AKC registered breeds. The Whatley Review is an online spoof newspaper as their fine print states.

Why didn't you include all 152 AKC recognized breeds?
If samples were provided to us by our deadline, we included that breed. Some breed clubs have not responded to our requests yet, or sent their samples in too late to be included in this particular paper. We're happy to include your breed in our coming studies. So if your breed was not represented, and you wish to participate by providing cheek swab samples from your dog, please contact dog_genome@mail.nih.gov and we will send you cheek swab kits. To analyze breeds beyond those included in our Science paper, we need at least 5 unrelated (to the grandparent level) dogs to give us a good representation of that breed's "fingerprint".

Are you still collecting samples? How can my breed participate?
A list of breeds that we especially would like to ask for donations is available by clicking the list of breeds link. Check the list to see if your breed is included. Contact us at dog_genome@mail.nih.gov to inquire about donating samples.

Some breeds have multiple heritages associated with the breed. Does this study distinguish the different lineages of breeds, such as reported in Akitas, Tibetian Terriers, or Manchestser Terriers?
No. Our study did not address this.

How does this study impact the health of purebred dogs?
The original motivation for the study was our desire to understand what breeds were closely related, so we can focus our attention on a subset of breeds when studying any given disease. So for instance, if we were to study a disease that is present in the Mastiff, we would hypothesize that same disease in the Bull Mastiff, Boxer, Bulldog, Miniature Bull Terrier and Rottweiler might all occur because of the same ancient mutation that each of these breeds carries. Studying multiple breeds simultaneously gives us much greater statistical power for finding any particular gene.

Other questions?
Please contact the AKC-Canine Health Foundation or us at dog_genome@mail.nih.gov.

DogHome