How To Decode Your Dog’s DNA

IRA FLATOW: This is Science Friday. I’m Ira Flatow. While we sit at home for months trying to entertain or improve ourselves, some people have been investing in finding out more about their furry quarantine friend by getting a doggy DNA test. Science Friday Producer Katie Feather looked into the new trend.

KATIE FEATHER: The first time I heard about consumer DNA tests for dogs, it was from my sister who got one for her dog Appa. Then my other sister found out about it and got both her dogs tested. A week later, I overheard my two neighbors discussing their dog DNA tests.

Sitting at home during the pandemic, staring at their pets faces, people had suddenly gotten very interested in doggy ancestry. Now, I’m a cat person. So I was tempted to just roll my eyes and say, this is just something dog people do. But as a science journalist, I was also curious.

What can these tests tell you about your dog’s genetic history? Are they the key to unlocking your pups behavior? Well, it turns out there are some experts in pet genomics, a term I didn’t realize existed, until very recently, who can help me answer these questions, and they join me now.

Dr. Eleanor Karlsson is an associate professor at the University of Massachusetts Medical School and the Broad Institute of MIT and Harvard. And Dr. Lisa Moses is a veterinarian, and research fellow, and instructor for Harvard Medical Schools Center for Bioethics. Welcome both of you to Science Friday.

LISA MOSES: Hello.

ELEANOR KARLSSON: Hi, thank you.

KATIE FEATHER: So how did you guys become the experts in these direct to consumer dog DNA tests? Because whenever I see them mentioned online, you two are some of the people who are most quoted.

ELEANOR KARLSSON: I’ve been working in dog genomics since I was doing my PhD, but it was many years later. And I met Lisa, and I can’t remember exactly where we were. But we got into a discussion about dog genetics and her perspective on it, which was very different from the perspective that I had as a scientist.

LISA MOSES: Part of the story of us getting interested in this is a realization that there was a big gulf between the science end of it and the clinical application of it. And I think we were both pretty surprised when we realized the ways in which these tests were being used and the way that they were being done.

KATIE FEATHER: Lisa, can you tell me, how are these dog DNA tests different in the clinical setting and these consumer directed tests, which are the tests that my friends or family are getting?

LISA MOSES: Well, there is definitely a big difference, and I think a lot of people who are doing the direct to consumer tests are getting back results with pages of information about potentially their dog’s future health. Part of the reason that Eleanor and I started talking about this is because we realized that people were getting lots and lots of information and really didn’t know what to do with it. I as a practicing veterinarian also didn’t know what to do with it. And as it turns out, once I started talking to Eleanor, there was good reason for me to not know what to do with it.

ELEANOR KARLSSON: A lot of people think that we understand genetics a lot better than we do. So when we do a scientific study, like the ones I was talking about, we tend to do a study, where we get a bunch of dogs, and we look for differences between the ones that have, for example, white coat color and the ones who don’t have white coat color and find things in their DNA that correlate with that trait. And it’s really interesting scientifically when what you want to answer is where does coat color come from, what could change in a gene that would make an individual have white fur as opposed to not have white fur.

But that’s a very different kind of question than what a lot of people think we can answer with DNA, which is making a prediction about what’s going to happen within an individual dog. And the problem is that an individual dog has that gene, but they also have 20,000 other genes. All of those genes interact with one another to create this very unique individual dog.

KATIE FEATHER: And we know from previous reporting, even on our show about these 23 and Me tests, that there are some ethical considerations. And Lisa, I imagine that that’s your role here is to kind of bring up some of the ethical considerations with these dog DNA tests. Are they the same types of considerations? What are you concerned about?

LISA MOSES: There are some considerations that are very much the same as they are for people, but some of them are different. The ones that are different that we’re particularly worried about is, first of all, because this kind of testing is really unregulated, we don’t know anything we’re not allowed to know, unless the company chooses to tell us how the tests are performed, what kind of technology they’re using. And most importantly, we don’t really know what the results mean. So when I say that, what I’m talking about is me as a veterinarian. If you come to me with your pet DNA test results that say that your dog might have an increased risk of having a particular kind of kidney disease, let’s say, we don’t have the kind of research to tell us how likely it is that test means that your individual dog is going to get sick, how sick, how likely is that overall.

We really don’t know any of that information. In people, before you can say that a particular change in the genes leads for sure to a disease, there’s a very high bar of information that has to be collected through lots and lots of statistically significant testing before you can make that claim. We don’t have that information in dogs. So it’s really impossible for a veterinarian to look at those results and say, yes, this is something you need to worry about. We need to do lots more testing, or no, you don’t need to worry at all.

KATIE FEATHER: A person might make a decision about their health based on their genetic test that would be concerning, like if they had an increased likelihood of having breast cancer. They might get a preventative mastectomy that may or may not be actually needed if they didn’t consult with a doctor first. And you mentioned that these same kinds of decisions could be being made by a pet owner, although, there is the added concern of whether a pet owner would want to put down their animal based on these tests.

LISA MOSES: For sure, and obviously, that’s the scariest part about this. One of the things that, I think, people really don’t understand the difference between a genetic test and a regular test that gets done at your veterinarian’s office is a genetic test doesn’t necessarily tell you that you have a disease, or you don’t have a disease. There’s a lot more to it than that, so it’s not like finding the parasite in the blood. Unfortunately, this makes it really tricky, because I’m worried that people will interpret the results in ways that they’re not meant to be used. I’m worried that people don’t understand that we don’t have all the follow up research that needs to be done to tell us whether individual diseases that are really connected to the DNA changes that are in the tests.

KATIE FEATHER: We’ve kind of covered what these tests don’t show you genetically about your dog, but let’s talk about what they are purporting to tell you. Eleanor, when a test comes back from one of these companies, typically, it’s going to report back what it says your dogs breed is, right? So how accurate is that information? Can you really feel good about the results that come back that is accurate?

ELEANOR KARLSSON: On a scientific level, this is something that we can definitely answer. So if a dog has ancestry from particular breeds, then we can compare their DNA to DNA from dogs from that breed and figure out how close a match we have there. I don’t know for sure how well the tests from the various companies work, because as far as I know, they’ve never kind of published in peer reviewed journals reporting on the accuracy of what their predictions are, which as a scientist is what I would like to see. But we do have algorithms that we use in our lab for figuring out what breeds a dog has ancestry from, and I can tell you that they work pretty well as long as that dog actually is a mix of breeds and we have information about those breeds in our database.

KATIE FEATHER: Yeah, tell me more about these breed mixes, because you were saying that most dogs are actually not any type of breed. Clarify what actually a breed is and how we use them.

ELEANOR KARLSSON: Yeah, so this is actually a really interesting idea that I’ve really just started thinking about a lot myself over the last few years. In the United States, as far as we know, most dogs are either purebred dogs, or they’re mixes of breeds. So that’s kind of the assumption that I made, and a lot of the very early genetic studies were all done with purebred dogs because of certain features of their genetics that made them easier to study.

But I was talking to somebody that studied dog behavior, a woman in my research group called Catherine Lord, and she said that, actually, most of the dogs on this planet don’t have any breed ancestry. And I went, wait a minute, what are you talking about? Then I thought about it a bit more, and one of the things we had figured out in our genetic studies was that most of the modern breeds are probably not more than a few hundred years old.

So if you kind of think about it, you’re like, how would all the dogs on this planet considering that dogs were probably domesticated like 15,000 years ago, they’re not all going to have ancestry from these breeds that are only a few hundred years old and only came from a few places on the planet? And it turns out that most dogs are just dogs. They’re not from breeds. They’re completely domesticated. They’re not any different from breeds or mixes. They’re just dogs.

KATIE FEATHER: So yeah, I was going to say, I think a lot of people’s incentive for doing one of these tests is to understand more about why their dog does something peculiar, or weird, or just to understand more about their behavior. So that they can accommodate them. So can they read something into their dog’s behavior when they get their test back?

ELEANOR KARLSSON: There are probably behavioral differences between breeds. There’s certainly been a lot of differences described, but they’re not as big as I was expecting them to be. You know, you kind of think about breeds, and people take a breed. And they’re like, golden retrievers– I don’t know. –like to retrieve balls. So then they think all golden retrievers are going to retrieve balls, but that’s just never true. Because dogs are complicated individuals. So while golden retrievers might be more inclined to retrieve balls, you’re going to have a lot of golden retrievers out there that really don’t care about balls at all.

KATIE FEATHER: So there are these direct to consumer companies set up to test your dog’s DNA, but you actually have a citizen science project in the works right now that does essentially the same thing. And it’s to help look at these patterns of behavior and genetics in dogs.

ELEANOR KARLSSON: Yeah, no, exactly. It’s a really interesting project. So when people think about genetics, they always think about the hard part is being this fancy technology to look at the DNA and do all these statistical tests. And it turns out that is not the hard part at all.

The hard part is finding out about the individuals. It’s finding out not their DNA, because you can get their DNA out of a saliva sample. And then you’re done. But finding out about a dog’s behavior, what they like to eat, you know, all of this kind of stuff is really complicated.

And for most animals out there that we want to study, you have to send scientists out into the field to find the animals, and watch them, and all this stuff. But the interesting thing with pets, including dogs and cats, is that every single one of them lives with a person or often multiple people that spend a lot of time watching what they do. So the way we set up our citizen science project is that you go online, and you sign up your dog. And then we ask you a bunch of questions about what your dog looks like and how they behave.

Then you can sign up to get a DNA kit, and we’ll sequence your dog. It costs about $150, and that’s done at cost. So it’s a nonprofit. The reason we want to have so much data is because studying the connections between something as complicated as behavior, or something as complicated as a disease test, or anything like that is that you need really big sample sizes to actually be able to figure out what’s going on to actually understand how much this change in your DNA is going to affect your health or your behavior.

LISA MOSES: And I have to say that, if you are a participant in that study, you get a chance to think about a lot of really fun things about your dog, like whether or not they kick after they poop and run around, and those kinds of questions that every dog owner can answer.

KATIE FEATHER: I’m Katie Feather. This is Science Friday for WNYC Studios. Lisa and Eleanor, we were e-mailing back and forth a little bit about my sister’s dog. Because I brought that up to you during our conversations before this interview. She had done one of these dog DNA tests, and her dog came back as 100% American Staffordshire terrier. I was just curious how likely this was to be true and how frequently a test would come back 100% of any type of breed.

ELEANOR KARLSSON: The percentages you come out with depend on how many different positions in the DNA you’re actually testing in the dog, and then, also, the algorithms that you use to analyze that data. So in the Darwin’s Art Project, we’re using a huge number of positions in the DNA that we’re actually testing, something like 15 million. Because our actual goal is to understand the origins of things, like behavior and health. And we need a lot of information about each dog to be able to do that.

So early in our study, we decided we needed to make sure we had all the 100 most common breeds in the United States in our reference panel. So this is the panel of dogs that we compare each dog to figure out what they have ancestry from. So if we don’t have a breed in that panel, then we can’t detect it.

So we went to a website, found out what the 100 most popular breeds were, and made sure we had them all. It turned out that was a list that had been put together by an organization called the American Kennel Club, and the American pit bull terrier is not part of the American Kennel Club. So it hadn’t been on that list. Anyway, as a result of all of that, we have dogs that we called before we put the American pit bull terrier in there and then after we put the American pit bull terrier in there.

And what we discovered was that, if we had a dog that have a lot of American pit bull terrier ancestry, and my sister’s dog Besko is actually one of these, and we didn’t have that breed in the reference data set, we tended to see that somewhere between like a half and 75% of that ancestry would get called a Staffordshire terrier. I’m guessing it’s because those two breeds share some ancestry. But it could be that, if you go to a company, they have a different way of analyzing the data, so that’s why it’s coming out at 100%.

KATIE FEATHER: Yeah, that’s just a good piece of information to have. It all depends on your reference genome, the reference that. If there isn’t that breed in there, then it has to go to the next best thing it seems like. What do you want people to know about these direct to consumer DNA genetic tests for dogs, and how do you, in general, feel about them?

LISA MOSES: Well, I will tell you from the veterinary ethics side of it, what I want people to know is please don’t use these tests to make medical decisions for your dogs. They’re just not ready for that yet. My hope is that with data sharing and large amounts of information collected from projects, like Eleanor’s Darwin’s Art Project, and hopefully, from some of the companies that will share their data, that we can learn a lot in the future about how dogs get sick, what the genetic component of that is, and that we can do something about it in the future. But I don’t think we’re ready to use the test to help your dog right now.

ELEANOR KARLSSON: I would just add to that. I think it is absolutely fantastic that people are interested in genetics and that they want to find out more about their dog’s genetics. Being curious is always a wonderful thing from a scientist’s kind of perspective.

You have to remember what it is the tests are telling you. They’re telling you that your dog has ancestry from particular breeds, but that doesn’t tell you what your dog is like. And it doesn’t tell you what your dog’s personality is. So as long as people can keep paying attention to the dog that’s sitting in front of them and not to what various books and websites are telling them, I think it could be a lot of fun.

KATIE FEATHER: I love that. Focus on the pet that’s right in front of you. So we’ve run out of time, but thank you so much to both of you for joining us. Dr. Eleanor Karlsson is an associate professor at the University of Massachusetts Medical School and the Broad Institute of MIT and Harvard. And Dr. Lisa Moses is a veterinarian, and research fellow, and instructor for Harvard Medical Schools Center for Bioethics. For Science Friday, I’m Katie Feather.

Copyright © 2020 Science Friday Initiative. All rights reserved. Science Friday transcripts are produced on a tight deadline by 3Play Media. Fidelity to the original aired/published audio or video file might vary, and text might be updated or amended in the future. For the authoritative record of Science Friday’s programming, please visit the original aired/published recording. For terms of use and more information, visit our policies pages at http://www.sciencefriday.com/about/policies/