Forum Replies Created
-
AuthorPosts
-
kathleen morrillKeymasterHi Jasmin,
Thank you for notifying us about the missing field for New Guinea Singing Dog in the feedback section — I’ll see that this gets updated, as it will be important for recruiting enough dogs to construct and update the breed reference panel.
Right now, we have 5 published samples from NGSDs, as well as 4 other NGSDs enrolled in Darwin’s Dogs (in addition to your dogs).
We aim to have 12 dogs to complete the panel for NGSD, even if they’re descended from a single pair, so that we can sufficiently cover any genetic variation that established throughout the population’s history. Your dogs will help us reach this goal, so thank you very much for your participation.
Sincerely,
Kathleen
kathleen morrillKeymasterWe may be able to add Carolina Dog in a future version of the breed reference panel as we acquire more samples.
As with all reference breeds, we need to gather 12 confident Carolina Dogs — a little tricky given that they aren’t “purebred” in the same sense as other breeds. We want to be certain the dogs in the panel are from the population they represent, rather than a group of unrelated mixed breed dogs, which might just add noise to ancestry results.
From publicly released DNA data, there are only 4 Carolina Dogs. In our database, we do have a lot of dogs indicated as Carolina Dogs — 29 dogs in total — but only three indicated as registered. Of those three, we have 2 samples on-hand, though not-yet-sequenced. Those that are sequenced show a high unknown percentage in their ancestry — I’d like to reach out to their owners to see what they know about their dogs’ backgrounds.
kathleen morrillKeymasterScience Magazine ran an article about the study this week, which comes out of the Arizona Canine Cognition Center. The manuscript is publicly available to read, too.
That’s correct — the study looks at C-BARQ personality surveys collected from many purebred dogs and corresponding genetic data averaged from different dogs (while still accounting for within-breed variation) rather than at genetic and personality data from the same individuals.
The results will be interesting to compare with our research in the context of some key differences, namely: different surveys to assess personality, including mixed and purebred dogs, using individual-level data, and considering breed ancestry.
kathleen morrillKeymasterHi Kate,
Yes, all participant results will be updated when the new reference panel is released! American Pit Bull Terrier ancestry should then be detectable.
kathleen morrillKeymasterHi Mary,
I’m sorry for the delay in adding American Pit Bull Terrier to the reference panel. We completed sequencing the 9 APBT candidate samples but discovered that 2 of them were mixed breed, not purebred. We have three confidently purebred APBTs being sequenced right now, so that we can complete the reference panel for 12 dogs in the breed. Then, I’ll be performing some quality checks and simulations to ensure the new reference panel is reliable.
The large percentage of unknown in Izzy’s current results do seem consistent with APBT not yet being detected.
Izzy’s and all participants results will definitely be updated when the reference panel updates. We will also send out an email notifying the update and which breeds we’ve added.
Thank you,
Kathleen
kathleen morrillKeymasterHi Holly,
Yes — so, the plot only shows breeds reported for at least 50 dogs swabbed, and Great Danes are a little less common in Darwin’s Dogs. There are 110 Great Danes enrolled, and 27 have been swabbed so far!
kathleen morrillKeymasterHi Julielani,
Jennifer is correct — through phasing the maternal and paternal DNA, we can achieve reach to the parental generation. Even then, phasing isn’t error free and can be difficult if the parents are similar genetically.
The process is more akin to human ancestry testing, which compares to DNA from representative populations. For example, if someone tests as 20% Italian, then that means 20% of their DNA is similar to modern Italians. The Italians in that reference panel may or may not be relatives of the person testing their DNA.
Generation/lineage tracing is best achieved when samples from multiple relatives within a family are compared. Popular human DNA testing companies have this feature because a customer usually has some family members tested as well. The best results come from also testing parents and grandparents.
Our breed results, therefore, might not fully reflect the reality of a dog’s family history. Most dogs don’t have a known pedigree, nor samples available from multiple members of their family. But we do sometimes get enrolled dogs within a family (siblings or parent(s) with pup(s)), so we may be able to examine those families to improve phasing.
One of our partner projects, The Working Dog Project, might achieve lineage analysis, as many working dog organizations have breeding programs in which they track family history.
kathleen morrillKeymasterHi Kelly,
I checked out your dogs’ results and feedback, and they do appear to be the correct DNA samples. Here are some possibilities for why the percentages and breeds may not be as expected:
For Beanie, you are correct that the Dachshund / Chihuahua ratio isn’t showing as a perfect 50/50. Do you know whether both parents are purebred Dachshund and purebred Chihuahua? Chihuahua can be a fairly diverse breed, so his parent may be a Chihuahua from a rarer lineage.
In Ellie, there are a lot of breeds detected at a low percentage. Ellie’s Beagle ancestry, even at 5%, could be shining through in her appearance and masking other breeds. It’s also possible that the greater percentage of Ellie’s unknown ancestry may come from a breed that is not yet in our reference panel.
If you have information on Brodie’s background and family history, feel free to share it through the feedback form on his page, too. I’m curious where the Corgi comes into his history. We do have only Pembroke Welsh Corgi in the panel, but this tends to be detected even in the Cardigans.
Thank you,
Kathleen
kathleen morrillKeymasterHi Mary,
I can assure you that American Pit Bull Terrier will be added to the panel no later than December. The reason that it hadn’t been included in the panel initially is because there was no verified, publicly-available genotype data from purebred APBTs. Our algorithm requires at least 12 dogs per breed. We already have data from 3 registered ABPTs, so we reached out to get samples from an additional 9 registered APBTs. These samples are currently being sequenced.
I can also confirm that Izzy is in the same sequencing batch as these 9 APBTs. So, the timing is such that there should be little delay between the return of her data and the updated breed reference with American Pit Bull Terrier!
APBT is a top priority for addition to the breed reference. Soon to be added, also, are Chinook, Saluki, and English Shepherd.
kathleen morrillKeymasterFor all of a dog’s DNA, there are two copies: one from mom, one from dad. Sometimes, these copies are different in code (A’s, T’s, C’s and G’s). So, this number is what percentage of the dog’s DNA code differs between copies. If the parents are very similar in their DNA, as with purebreds, then this percentage will be low.
kathleen morrillKeymasterMost purebred dogs fall between 10% and 25% (usually 20%) and most mixed breed dogs have between 25% and 30% (although I’ve seen a few with genetic diversity as high as 40%!)
kathleen morrillKeymasterHi Casey,
In short, the unknown percentage of your dog’s ancestry might not be just one big section of mystery in their DNA… but many, many small sections of mysterious DNA that add up! It’s not clear what or how many breeds may exist in the unknown.
One way we can find out is when an owner gives us feedback that they know their dog is a certain breed not yet in our reference panel. These dogs come up with a high percentage of unknown, or a closely related breed. We’ve seen this with American Pit Bull Terriers: high percentage unknown, high percentage Staffordshire. This was a good sign that we needed to add APBTs to our breed reference panel.
We expect to minimize much of the unknown percentage as we improve our reference panel by adding more breeds.
kathleen morrillKeymasterHi Becky,
The genetic diversity value we measure is what percentage of genetic markers do the maternal and paternal DNA differ. Most purebred dogs are genetically similar within a breed, and we can use this to draw a decent line for “likely purebred”.
This isn’t a very robust measurement of relatedness in their family tree, compared to COI described here. Some genetic markers rarely ever differ across dogs — related or not — and so a better measurement won’t treat all markers equally. We are currently working on achieving a better genetic diversity measurement, so you can expect this to improve with time.
kathleen morrillKeymasterHappy birthday, Pippi PugStocking!
Seeing as Pippi is 94.8% Pug and certainty for that match is 97.1%, we are confident she’s a Pug! We’d likely detect other breeds in her results at a higher percentage if they had been mixed into her pedigree recently.
Her looks may just be within the normal variation of Pugs. I saw a post recently on a dog DNA testing forum about a Pug like yours — longer nose, longer body, and taller — who also tested 100% Pug.
kathleen morrillKeymasterHi Lindsay,
Insightful question — that is exactly what we do: both old and new answers are saved with a timestamp. Another student and I will soon be mining survey data to determine the age (or approximate age) of dogs when their owner gave a survey response. This will be quite important for surveys that relate to canine age-related cognitive decline.
We have more plans to track changes in behavior over time, such as encouraging owners who register puppies to return for new surveys as their dog ages.
-
AuthorPosts
