[Jennifer]

It sounds like maybe you’re describing scent rolling? That’s a wolf behavior which many dogs retain. Our mutt is an extremely enthusiastic scent roller as well. My understanding is that it’s really not all that clear from the existing research what purposes it serves in wolves. The most common theory seems to be that it may be a way to carry important scent discoveries back to the pack (food, rivals, dangers, mysterious novelties etc.), although wolves will scent roll when the whole pack is together too, and apparently individual wolves also vary quite a bit as to which scents they find worthy of rolling in, so that may not be a fully adequate explanation. Anyway, there are lots of videos out there of both dogs and wolves scent rolling, so you could check those to confirm if that’s what Taz is doing.

[Jennifer]

Not Mini Schnaus, but purported/presumed APBT x Jack mixes aren’t uncommon at the city shelter I volunteer with, and I’d guess the temperament would probably be fairly similar? Both APBTs and Jacks are very common around here, so I assume most of these mixes are one-off oopsies, since it seems unlikely there’d be much of a market for pit x terrier crosses specifically. From what I’ve seen they’re usually a bit more easygoing than a pure Jack, but with just as much prey drive and a bit more tendency to be dog-aggressive. Game-bred APBTs can run pretty small; 30ish-40ish lbs. isn’t unusual here. Mini Schnaus aren’t quite as popular as they used to be–AKC top 20 now, down from top 10 in the late 20th century–although that probably varies widely by area; I seldom see them locally. I assume your dog has the wiry/bearded coat, always thought that looks cute on pit mixes.

[Jennifer]

I have no Darwin’s Ark dog links to provide, but have you seen Embark’s page about American Village Dogs? It has links to the full Embark results for a dozen randomly selected dogs who’ve tested as 100% AVD with them, as well as for a dozen AVD mixes they’ve tested.

https://my.embarkvet.com/breeds/american_village_dog#dogs

I sometimes browse a subreddit devoted to sharing DNA breed test results, and “rez dogs” posted in there not infrequently test as partly or entirely AVD, which as you say makes sense. Others test as multimixes of various breeds, or as mostly or entirely “Supermutt” (i.e. matches neither any known village dog populations they have on file, nor any pure breeds on file, and therefore probably does descend from purebreds but it’s too far back for reliable matches).

[Jennifer]

I don’t know what Nigel looks like, but with reference to his current breed results as mentioned in your other thread, I could definitely imagine a “Chorkie” coming out looking a fair amount like a very petite Lancashire, especially if its Chi parent were short-haired (dominant trait). All three of his primary breeds have prick ears, and they can also all be black-and-tan like Lancashires (recessive). Being 20 lbs. and dwarfed (I’m assuming you’re saying he’s dwarfed i.e. ultra-short legs) probably couldn’t be explained by those three breeds alone, but the other 10% might include a dwarfed, much heavier breed(s). In any case, Lancashires are pretty rare, even in the UK, so for a shelter mutt to be high-content-Lancashire would be very surprising, although stranger things have happened.

With this latest revision, my dog also showed gains of % ancestry for 4 of the breeds already detected in her. In her case this probably(?) has something to do with much of her ancestry consisting of rather closely related herding breeds, therefore lots of confounding genetic overlap. I’m not sure whether something like that could apply to Nigel. I think it’s maybe kinda like trying to piece together an enormous jigsaw puzzle when you don’t know what the final image is supposed to show…those portions of it that initially fall into place for you fairly quickly, you’ll probably still have to revise later, once you’ve figured out what some of the other components are, because your working mental image of each component is continuously getting modified by new information about other portions. Even so, assuming the person doing the puzzle is reasonably mentally competent, their initial misconceptions are still going to fall within a certain limited range.

[Jennifer]

I’m not sure if this helps, but as someone who has experience with purebred ACDs, I just wanted to add that the “natural suspicion of strangers” the ACD breed standard calls for is very different from “stranger danger,” which fearful dogs of just about any breed can show. The ACD standard is referring to a certain coolly confident, calm, sharply scrutinizing “tough guy” stare that ACDs typically show when encountering strangers, which puts the stranger on notice without being dramatic. (By contrast most Border Collies show strangers polite reserve, without that air of sharpness.) But if instead a dog shows a tendency when encountering strangers to snap directly into a full-tilt pre-emptive threat display (nervous growling and tensing, reactive lunging and barking etc.), that’s “just” fearfulness, and can’t safely be assumed to have anything to do with any one underlying “natural” disposition towards strangers.

Because the working herding breeds tend to overlap so much in behavioral traits–extremely biddable and handler-oriented, great all-around natural athletes, obsessed with controlling movement, standoffish with strangers and needing lots of socialization in puppyhood to be comfortable around them–it can be really tough to pinpoint breed-specific behaviors in a herding mix of uncertain ancestry. Sometimes the tactics the dog favors during herding play can be a strong clue (e.g., dropping into a stalk and giving eye vs. darting in close to grip the heel), but even this isn’t foolproof, because none of these tactics are truly breed-specific, and besides a dog doesn’t necessarily need a lot of ancestry from any one breed to show its stereotypical herding behaviors.

Anecdotally BCs do have the reputation of being the most prone of all the herding breeds to fearfulness problems, but I’m not sure what actual veterinary behaviorist data might have to say about that, let alone the genetic factors that Darwin’s Ark is looking into.

[Jennifer]

So you’re on St. Kitts then? That must be an interesting place to be a vet! For some reason, the image links for your dogs aren’t working for me (I think you have to either link to photos in your DA profile, or else post an external link to Imgur or similar). But anyway, I did look up some pics of “St. Kitts brown dogs” and also Googled out of curiosity to see if Embark had any test results for such dogs publicly available, since they test for “American Village Dog” as well as for mixed breeds. They did have some:

Link One
Link Two
Link Three
Link Four

I did also notice results for a couple dogs from St. Kitts who tested as, and looked like, predominantly-pit-type (AmStaff/APBT) multimixes, so that would be more like your dogs, presumably. I’m not sure whether Embark also maintains a disclaimer regarding dogs from outside certain countries. Their “village dog” reference panel dogs are definitely from all over the world, because the founders of the company already had that DNA from their own research, but that doesn’t necessarily mean their reference panel dogs for the pure breeds they test for are comparably internationally diverse.

————————————————————————————-

I’ve had my mutt (the dog in my avatar) tested with both WP and Embark. She was born in a shelter, and the shelter people had seen her presumed sire as well as, of course, her dam, who were stray traveling companions. They guessed that mom was either a high-content-ACD herding cross, or else an unusually lanky pure ACD, and described dad as looking “like a Boxer/Collie mix.” Within a short time after bringing Luna home, we were pretty sure the ACD and Boxer parts were correct (nip nip nip, boing boing boing). But we were really curious to see what the rest might be, because her build is distinctly more delicate than a 50/50 ACD/Boxer’s would generally be, and she also showed some behaviors that, while herder-y, didn’t seem quite ACD-ish (very “soft” and almost comically earnest about training, flashes of “strong eye” behavior during herding play).

her WP results:

~ 50% Australian Cattle Dog (predicted as one purebred parent)
~ 25% Boxer
~ 12.5% Australian Shepherd (asterisked, indicating uncertainty as to whether the predicted % is really accurate)
~ 12.5% too mixed to call/unknown, with Herding and Sighthound groups listed as possibilities

her Embark results:

34.9% ACD
15.1% Border Collie (predicted to come from the same parent as the ACD, therefore one “Border Heeler” parent)
25.1% Boxer
15.8% Australian Shepherd
9.1% too mixed to call/unknown, with Sheltie specifically predicted to be in the mix

her Darwin’s Ark results to date, by % of ancestry (hopefully these will fill out more as their reference panel grows):

35.6% ACD
26.7% unknown
21.2% Boxer
6.2% Border Collie
3.8% Collie
3.6% Sheltie
2.9% Australian Shepherd

So overall, pretty similar results. And all three work pretty well to explain her looks and behavior, I think. Both WP and Embark were pretty much spot-on about her weight, but from what I’ve seen that’s often not the case, even when the reported mix itself seems very probable.

[Jennifer]

Thanks Kathleen. I guess I may change my answer re 3) then, because Luna definitely has sable tipping showing in her body patches, I just wasn’t sure whether that “counted” or not.

Re 9) it’s definitely not common, but I’ve volunteered with our local public shelter for a couple decades, where we get tons of pit mixes, and I’ve seen it maybe a few times in bluenose pits (not puppies). Usually they turn that yellowish-greenish-amberish color but apparently not quite 100% of the time.

[Jennifer]

It seems like most of the genetic tests commercially available nowadays, whether from UC Davis or Embark or whatever, will at least give you a rundown of your dog’s genotype for several of the genes presently known to be associated with certain physical traits–coat pattern and texture, overall body size (numerous genes involved in that) etc., in addition to level of inbreeding. And Darwin’s Ark is looking at much more of the genome than any of those tests, so I’d imagine(?) all of that info and more would be available. How “useful” that info is would depend on your interests–like, I’d be curious to know if my dog might be merle (she has very little black coat pigment, so I can’t reliably tell by looking) because if she were it might explain some odd quirks about her coat patterning and eye color, but I don’t know that that’d really be “useful” info so much as personally interesting to me. But what’s really exciting to me about this study is not so much anything I’d learn about Luna specifically–I already know her behavioral quirks, and as far as it goes no DNA test is going to help me manage those–but rather the opportunity to contribute in some small way to furthering scientists’ understanding of the genetic bases of dog behavior, which in the long run could be very helpful for breeders too.

[Jennifer]

Do Bee’s Embark “Family Tree” results correspond fully to her “Summary” bar graph? I.e., does the tree show 4 great-grandparents labeled “Golden Retriever” and 4 labeled “Bloodhound?” Or is one or more of them possibly labeled “Golden Retriever mix” or “Bloodhound mix?” Because sometimes Embark tweaks those Summary bar graph numbers a bit to make them add up to 100, even when the family tree suggests otherwise. This happened with my dog, actually–the Embark breed %s I’m quoting for her are taken from her Summary, but her Summary actually just says “9.1% Shetland Sheepdog,” whereas on her family tree, one great-grandparent is labeled “Shetland Sheepdog mix,” which confused me because all the other breeds they’d called were already represented on her tree as purebred GGs, and the Summary numbers added up to 100, implying that all DNA had been accounted for. I emailed them and they said her family tree was more accurate in that respect, that in fact they were seeing some too-mixed-to-call DNA in association with her Sheltie DNA.

If that’s not the case for Bee though, then I got nuthin’.

[Jennifer]

My mutt has been tested with WP, Embark, and DA.

WP:

~ 50% Australian Cattle Dog
~ 25% Boxer
~ 12.5% Australian Shepherd (asterisked, indicating uncertainty as to whether the predicted % is really accurate)
~ 12.5% too mixed to call/unknown, with Sighthound and Herding groups listed as possibilities

Embark:

34.9% ACD
25.1% Boxer
15.8% Australian Shepherd
15.1% Border Collie
9.1% too mixed to call/unknown, with Sheltie specifically predicted to be in the mix

DA:

35.6% ACD
26.7% unknown
21.2% Boxer
6.2% Border Collie
3.8% Collie
3.6% Sheltie
2.9% Australian Shepherd

In my dog’s case, I’m guessing(?) the main reason for the differences may be that she really is roughly 75% assorted herders (as WP and Embark found), but that without both a very large number of markers and a very robust reference panel, it becomes difficult to accurately assign all her non-Boxer DNA to one-or-another herding breed, since all the herding breeds detected in her to date are thought to be members of the extended collie/UK sheepdog family (Parker et al. 2017), and therefore pretty closely related. So I’ll be interested to see how much of her current 26.7% unknown winds up getting reassigned to already-called herding breeds. Would be just as ready to believe there’s a splash of something that hasn’t been called yet knocking around in there, though! The ACD and Boxer have been pretty obvious in her from the beginning (nip nip nip, boing boing boing) so I’m not surprised those two findings have held fairly constant across all three tests.

[Jennifer]

I think you might be somewhat confusing the American Bully with the APBT? The APBT, AmStaff, Staffy Bull, Bull Terrier, and Mini Bull Terrier all originate from the same early-19th-century crosses in the UK (old English Bulldog x assorted terriers, likely with a pinch of Whippet as well). Over time, different strains of that basic “Bull-and-Terrier” type developed. The BT and MBT were the first to split off from the rest, in the mid-19th century (this is all still happening in the UK). Then the Staffy Bull and APBT diverged in the late 19th century, due to the lines which became the APBT having been taken to the US by immigrants. The AmStaff was developed later (in the US, early 20th century) out of various APBT strains, and is registered through the AKC whereas APBTs are registered through the UKC. (While AmStaffs and APBTs remain very closely related, AmStaffs tend to be somewhat taller, stockier, mellower, and less athletic–sort of like a show-line version of the APBT, whereas the APBT remains well-suited for sports, farm work such as hog-catching, and, unfortunately, underground dog-fighting as well.) Finally, the American Bully was developed in the late 20th century by crossing AmStaffs, APBTs, and assorted smaller bully-ish breeds–it’s not really clear which smaller breeds were used, but Frenchies and Staffy Bulls were almost certainly among them. (In fact, this is thought to be how Bostons originated as well, excpet they’ve been around since the late 19th century.) The AmBull was the last of these breeds to be recognized by a major kennel club (UKC, 2013).

However, the vast majority of “pit bulls,” including many who are marketed as supposed purebreds of one of the above breeds, are non-pedigreed dogs who very often have several of the above breeds, as well as pinches of various other breeds, in them (e.g., mastiff types for a bigger and tougher-looking “pit bull,” Bostons for a smaller and cuter-looking one, etc.). While there’s no way for a DNA test to tell which crosses in a mutt’s background were deliberate vs. which ones were “oopsies,” Lola’s results to date could fit well with her having either American Bully, or else some generic “pit bull” meant to mimic an American Bully, in the near reaches of her family tree. As for the Mini Schnauzer x probable(?)-multi-bully-mix at the parent level, that would almost certainly have been an “oops”–I doubt anyone is breeding that mix deliberately, as there wouldn’t seem to be much of a market for it.

[Jennifer]

My mutt’s reported genetic diversity is 22%; is that unremarkable for a multimix with completely unrelated parents? Embark reported her CoI as 0%, fwiw. She does have some rather closely related herding breeds between the two sides of her family, so I could see there being some increased overlap because of that.

[Jennifer]

One thing I’ve noticed, in browsing a subreddit where people share their dogs’ DNA breed test results, is that people can sometimes be quite resistant to accepting a finding that their mutt contains only a very small percentage of “Breed X,” when said mutt happens to have inherited some particularly dramatically visible trait associated with “Breed X.” For example, a dog who has a merle coat pattern, or whose coat texture is wiry-scruffy with facial bearding. Both those traits are dominant or co-dominant and caused by just one gene, so even a tiny amount of ancestry from a merle or wiry-coated breed could be sufficient to cause them if the portion of DNA inherited from that ancestor happened to include that gene. But they’re also both very “loud” traits visually, and that loudness can trick us into assuming that there must be “a lot” of some merle or wiry-coated breed in there. And once you’ve gotten that idea into your head, it can be easy to then proceed to convince yourself that your mutt “clearly” also shows other, more vaguely defined and/or multi-gene traits associated with that breed, like “high energy” or “feisty” or “fast” or “leggy”–these are all pretty relative terms, and someone with years of experience in a breed commonly described as having them may understand them quite differently than the average person.

Also, even within most pure breeds, there’s sufficient diversity present in the collective gene pool to yield individuals who may be quite far off the norm or ideal for that breed (as epitomized by show or field champions) in both looks and behavior. Sometimes this gets read simply as that individual being a “faulty” example of the breed, other times it may lead to accusations or suspicions that the individual isn’t really purebred. There can be characteristic variations within lines, too; for example, German Shepherds bred for police work are as a group likely to display a somewhat different balance of behavioral and perhaps physical traits than those bred for seeing eye dog work, which in turn will look and act somewhat different than GSDs bred for show ring success. Backyard-bred and puppy mill dogs of a given breed, who generally haven’t been bred according to any selection criteria at all–just “whatever two dogs of this breed I happened to already own”–are especially likely to display variation from the breed standard, because standards are meant to define an ideal that breeders must keep continuously and rigorously selecting for; they aren’t meant to summarize all possible individuals that could result from the genetic diversity naturally present in the breed.

[Jennifer]

For the morphological predictions, since that’s not a focus of the study per se, would that basically take the form of noting which alleles the dog is carrying at genes already known to affect physical phenotype? (e.g. curly hair, brachycephaly, merle) Or do you also have ambitions to identify previously unknown causative mutations, or to share data with researchers who are studying physical phenotype? For example, I have a purely armchair interest in color and coat pattern genetics, and I know that even in that one small and admittedly-not-too-important-in-the-big-picture area, there are still so many unanswered questions despite the relative simplicity of identifying good candidates for study, probably largely due to a shortage of funding for research into the topic.

[Jennifer]

The breed testing process, including number of markers used, was discussed in the FAQs for the MuttMix project, a collaboration between Darwin’s Ark and IAABC:

http://iaabcprojects.org/faq/

“Our reference panel of dog breeds has information for about 200,000 markers in each dog, so these are the markers we currently use in our breeds test. Our technology, however, tests at least 1 million markers in each dog. For some dogs, we have information on all 2.4 billion nucleotides in their DNA. Over time, our breed calling algorithm will perform better and better as we change over to a reference panel that includes just as much information for each dog.”

For comparison, here’s Wisdom Panel’s summary of their process:

https://help.wisdompanel.com/s/article/How-does-Wisdom-Panel-work

“Once your sample is received at our lab it is scanned into our system and assigned to a batch for testing. It then undergoes processing to extract the DNA from your dog’s cells, which is examined for the 1800+ markers that are used in the tests. The results of these markers are sent to a computer that evaluates them using a proprietary algorithm designed to consider all of the pedigree trees that are possible in the last three generations of your dog’s ancestry. Our computer algorithm uses samples from our extensive breed database to analyze these potential pedigrees and calculate which one is the best match.”

…and Embark’s summary of their process:

[link deleted, because it created a huge obnoxious Embark logo…lol…but, this is from the FAQs on their website]

“At the lab we extract your pup’s DNA and run it on our custom-built genetics chip, which is a proprietary DNA microarray with over 200,000 markers…Ultimately, we use several proprietary algorithms to build the story of your dog and his or her DNA one chromosomal segment at a time. If numbers are your thing, we test about 256 quadrillion (that’s 500 million times 500 million) different possible genetic ancestry combinations for your dog using the latest and greatest science.”

So, Darwin’s Ark’s breed testing process is different from both WP’s and Embark’s in that they’re using low-coverage sequencing, which does test many more markers, but their reference panel of purebred DNA is still in the process of being built up, simultaneous with the DNA testing of dogs they’re already doing. (Initially, their reference panel contained only whatever breed-specific DNA information was publically available, as alluded to above.) So, owners of dogs who’ve already received their initial breed test results will most likely see those results update over time, with an increasingly lower % of “Unknown” ancestry as the reference panel becomes more robust.

All three tests use algorithms to analyze each dog’s results for their chosen markers against the purebred dogs in their reference panel. Wisdom Panel’s and Embark’s algorithms also produce three-generation predicted “family trees” for each dog, whereas Darwin’s Ark reports their breed calls as a simple bar chart ordering the detected breeds from lowest to highest percentage of ancestry.

But probably the biggest difference is that Wisdom Panel and Embark are for-profit companies focused specifically on breed analysis (Embark does also publish some research based on their findings), whereas Darwin’s Ark is a “citizen science” study focused on the genetic bases of dog behavior, with breed analysis effectively being a side benefit of your dog’s being selected for one-or-another phase of the ongoing study. And participating is free (unless you choose to donate, as explained in the FAQs, under “About” at the top of this page).

[Author]

Posts