Michelle Moyal, DVM (Host): Hello, everyone. Happy 2026 and welcome to the latest episode of the Cornell Veterinary Podcast. I'm your host, Dr. Michelle Moyal, Visiting Assisting Clinical Professor, Purina veterinarian. And might I add two additional things for this new year? One, new glasses wearer, which is very interesting, because I'm getting older. Let's just put that out there; and two, a big fan of unrivaled basketball, which is starting. So everyone, if you haven't been watching, start watching now.

But I'm very excited. You know how we do here. We deep dive into the discovery and the learning that happens at Cornell University's College of Veterinary Medicine. And today, we are going to jump right into a really fun topic. I'm very excited about it. With me is Dr. Adam Boyko. My hands are flying, everyone. If you're not watching this, you should be because that's how excited I am.

Dr. Adam Boyko, Associate Professor in Biomedical Sciences and Chief Science Officer at the dog genetic testing company, Embark Veterinary, which I hope most dog owners-- myself included; I'm waiting for test results, like, as we speak. We'll talk about that in a little while-- may be familiar with. So, welcome to the show.

Adam Boyko, MS, PhD: Thanks, Michelle. It's great to be here.

Host: I'm going to read your bio because I think it's one really cool, and I want people to know a little bit of your background. But then, I'll have you kind of expand upon it for sure, right? So as a researcher, you've studied the genomic evolution of dogs, and I want us to elaborate on that. You also investigate dogs as a model for genetic diseases. And you teach a summer course, or you will be, on these topics for our budding canine geneticists out there. Hello, everyone listening. Just you wait, you're going to be excited to become a geneticist in the future. You're a graduate of the University of Illinois at Urbana-Champaign. You received your master's in Computer Science-- oh, cool. I like to hear about different journeys into our science spaces here-- and a PhD in Biology from Purdue University before you did your post-doctoral work at Cornell and Stanford. Wow. You know, no big deal, just Cornell and Stanford. But welcome to the show officially, Dr. Boyko.

Adam Boyko, MS, PhD: Great to be here. Thanks.

Host: So, I always like to ask people their origin story, right? Because you have a degree in Computer Science and Biology. You could do a lot of things with that. So, was this your original final destination? Like, what were you thinking about? How did you get here?

Adam Boyko, MS, PhD: No, I mean, I always loved animals, although I'm allergic to cats and horses and things. So, I didn't think I would ever be a veterinarian.

Host: Yeah. Purina has a diet for that, by the way. If you're allergic to cats, Purina's got something for you. I just wanted to share. Okay.

Adam Boyko, MS, PhD: But I was really interested in science in high school. I had a great physics teacher, so I decided to start majoring in physics. And then, I took some college biology classes where I got to go out in the field and stuff. And I'm like, "Actually, this is way more interesting." And so, I did an undergrad in Ecology, Ethology and Evolution. But at the same time, I also got a degree in Computer Science, because I really liked coding, I liked building stuff, software kind of stuff. And so, I was a double major. And then, I decided, do I want to be a computer scientist that knows some biology or do I want to be a biologist, you know, some computer science?

So, I decided to get my PhD in Biology. But while I was there, got a master's in computer science. And at the time, a lot of the biological interest in computer programming was in like solving protein folding problems and things like that that didn't really interest me that much. But by the time I was graduating bioinformatics, the human genome, all of this genetic stuff was coming online. I'm like, "This is really, really interesting." My PhD work was on Heliconius butterflies, so that got me out into the tropics, exploring the rainforest. You know, I'm pretty good--

Host: Is that a certain species of butterfly for those of us butterfly illiterate?

Adam Boyko, MS, PhD: It is a genus of butterflies. It's the classic mimicry complex. So, you have all these different species with different color patterns in different forests and they're trying to avoid predation by the local birds and things like that. And so, it's kind of an interesting--

Host: No one wants to be eaten.

Adam Boyko, MS, PhD: Right, right. And so, I brought some of the butterflies back. I had my own colony at Purdue. I was doing spatially explicit simulation modeling for how different color patterns could evolve and stuff like that. And it's really neat because I'm still paying attention to the research in that area. And they've decoded the genomes. And now, they've figured out the genes that are underlying all of these different color patterns. And so, it's really kind of fascinating.

So, I got my PhD. And I was fortunate enough to come to Cornell to work on population genetics. And so, I was in a lab, it was mostly human genome data at the time. So, that's what I was working on, building mathematical models, ways of analyzing bioinformatic pipelines, things like that. And then, the lab happened to start collaborating on a dog project. And I kind of got put on the dog project because we needed to fix some of the bioinformatic pipeline, and then do some of the analyses. And I thought it was really cool. I was like, "Wow, dogs are almost like butterflies, and that they have these really cool patterns that have evolved out of nowhere," and the evolutionary basis of that. And it was a really, really fun collaboration. And we found out genes for floppy ears in dogs, the genes for body size in dogs, the genes for different coat types in dogs. It was the biggest dog genetic data set at the time. It was about 800 dogs we had from about 80 different breeds, including wolves. So, it was really fun.

But people were really interested in the origin of dogs. And for that, looking at pure-bred dogs is not the best way to figure out the dog origins, because we've really modified these pure-bred breeds. So, it's really great figuring out the genes that are different between tall dogs and short dogs. But it's not great for figuring out what are the original dogs.

Host: Right.

Adam Boyko, MS, PhD: So, Cornell gave me a seed grant to go traveling to find street dogs and village dogs around the world. So, we started out in Puerto Rico, then we did some in Africa. I worked with my brother on this because he was a master's student in Public Health. And so, he was going to all these places anyway. We ended up doing--

Host: What an awesome collaboration with family!

Adam Boyko, MS, PhD: It was really fun, because he loves field work. I love field work. We just had a great time traveling. I mean, we sometimes did solo, but we were in Peru together for a month. We did various islands in the South Pacific. We had National Geographic funding to do that. We got NSF stimulus funding. The project moved when the lab moved to Stanford. We continued doing that. We got whole genome sequences from dozens of dozens of these village dogs. We got a whole bunch of new insights, including that the Center of Diversity for dogs is actually in Central Asia. And so, that's probably where dogs originated.

Now, there's even bigger data sets out there with ancient DNA that are sort of corroborating where this center of diversity is and how many domestication events happened, and what that whole process was. And so, it was really nice to be able to contribute to that and have that project that I could bring to Cornell when I came back, when there was a faculty position open specifically for canine genetics. So, Cornell's always been kind of a leader in dog genomics. And so, it was really fun to come back working with the same colleagues I was working with when we left, and then being able to build up a lab and expand out the different kinds of questions we could ask with dogs.

Host: So, a few things that I've heard that I really like, and I find a lot of commonalities when people tell their stories here. And even though I blacked out-- and everybody will know this-- when you said mathematical modeling, I did hear some other things, and those things are important to me.

So, what's important to me is you found this interest in genetics and you were studying butterflies. And while you didn't seek out this dog project specifically, you were asked to help on it. And then, you found this passion that you didn't even know existed and you even said-- dare I say it, listeners-- that you had so much fun doing it. And then, you get this amazing opportunity by receiving some funds to travel and literally try to find the origin of where dogs came from.

And while I joke that the pug I see in practice maybe doesn't resemble the village dogs that you were studying, right? We have so many breeds now. I'm still a pug lover, everyone, right? We have all these breeds now, but they're still not-- right. We've done so much selective breeding, that wasn't the information you were seeking out. So, you had a question and you went to go find the answers. And that answer was in Asia. Wow. And then, here you are, full circle moment back to where they made you help them ask those questions back doing this work.

Adam Boyko, MS, PhD: Right. It was really a lot fun. And it was a big debate, because some people thought that dogs had originated in the Middle East because the cradle of agriculture, and there was genetic evidence that dogs were closest to Middle Eastern wolves. But as we've studied more on the genomes, including ancient genomes, that's because Middle Eastern wolves have admixed with dogs. That's why they're a little bit closer. So, they carry some dog DNA, so that was that signal. It wasn't that dogs came from the Middle East.

Then, there fossil evidence in Europe. But of course, the fossil record in Europe is a lot more complete than the fossil record elsewhere. And so, it's really needing to look at modern and ancient dogs and wolves to try to figure out where the diversity overlaps, and we could find an origin.

Host: Yeah. That's really neat. So, was it a very specific place in Asia? Did you narrow down--

Adam Boyko, MS, PhD: No, it's very broad. We have a pretty broad range right now. We're certainly interested in trying to-- is it Western China? Is it Mongolia? Is it closer to Nepal? All these places where we have some genomes that we really need to dive down a bit deeper and see if we can find ancient samples that might reveal with more granularity precisely where they came from.

Host: Yeah. Okay. So, that's very cool. So, that leads me to my next question. What do you think is like your favorite discovery thus far? Like, favorite findings related to genetics? Like why? What did you find? You were like, "Yes, This is it. I love this!"

Adam Boyko, MS, PhD: I think probably the very first unexpected finding that I got. So when we were working with 80 different breeds of dogs, we had, I don't know, like 50 different measurements on these dogs. Fur measurements for the breed, skeletal measurements for the breed. So, we found genes associated with body size, genes associated with limb length, genes associated with skull, snout length and things like that.

And we could take all of the genetic data we had, and I could run through this model and show regions of the genome that have been differentially selected in different breeds. And we could overlay those two. And we could see, "Okay, well, this is associated with body size and we see a signature selection there. This is associated with skull shape and we see a signature selection there." And the biggest signature selection that we saw in the whole genome didn't map to any trait that we measured. So, we had a group of dogs that looked like wolves in this part of the genome and a group of dogs that looked very different and we clearly selected for something different.

And, finally, I it wasn't a region that had really been associated with a pheno-- it was kind of close to a body size gene, but not right over it. So, I finally did like a Google search image for the different breeds in the different categories. And after doing about a dozen of them, I was like, "Oh, this is the gene for floppy ears." So, the breeders have to fix the ear type in each breed, and some breeds have prick ears and some breeds have folded or floppy ears. And so, that was the different-- between whether you have the wolf-like prick ears or a dog-like folded or floppy. So, we hadn't thought to measure that ahead of time before doing the study. And so, the things you come up with if you're willing to look for them are oftentimes not the things that you went into the study to figure out. So, that was really cool.

Host: Yeah, you're taking me back to when I was a kid reading like a Highlights Magazine. For those of you listening, a long time ago, we used to have pictures we looked at. And we had to find the differences and kind of circle them. I feel like that's what you were doing, just like trying to figure out the differences. That's so neat.

Okay. I'm going to ask a question that's not in my list of questions. Cornell loves when I go off script. Is there a breed of dog, domesticated breed of dog right now, that most closely resembles kind of our origin? I'm just curious. Do you happen to know?

Adam Boyko, MS, PhD: No, I don't think that there is, I think we've sort of modified all of these breeds, and some of them maybe haven't gone through as big of a bottleneck. If you think of your typical street dog around the world, I mean, there's a billion dogs in the world, most of them are not pure-bred dogs. Most of them are not mixed breed dogs. Most of them are just dogs that are wandering around human developed areas like they have been for millennia. And they tend to be medium sized, they tend to have short fur, they tend to be brownish, but they can be black. It's like the default dog. We just sort of call it a village dog.

Host: Wow. But there are defaults kind of. You see some default, like, details in these dogs. Like you were saying, medium size and coat color.

Adam Boyko, MS, PhD: Right, right. Yeah. But of course, it'll vary. So, we have corded coats on the dogs up in the Andes, right? And we see extra digits on the toes of some of the dogs in some islands and things like that. And you do have islands that have dogs that have short legs, like corgis--

Host: Yeah, I'm intrigued. I just need a list of all of these. Okay. Okay. I'm back on track. I'm getting back on track, because I really want to talk about Embark. And my first question to you is you basically found this company based on research you were doing. So, first of all, could you share with our listeners what Embark is in case they don't know, kind of how you decided to start it and like how it has grown into what it is today?

Adam Boyko, MS, PhD: Yeah. So, Embark is a dog DNA testing company. It started 10 years ago out of Cornell. So, we got what's called a Cornell Technology Grant to start kind of building out the platform. It was an offshoot IP that was developed in my lab and elsewhere at the vet school.

Host: What's IP? So sorry.

Adam Boyko, MS, PhD: Intellectual property.

Host: Intellectual property. Got it.

Adam Boyko, MS, PhD: So, we had figured genes that do different things, working out algorithms for improving the way we build the genetic platforms that we use, and then interpret that to give results to potential customers, right? And by doing the dog research and having the research covered in scientific journals and the popular press, there was lots of inbound interest in, "Can my dog contribute to your research? And can you tell me more about my dog's genes?" right? And so, it was this idea that dog DNA testing at the time was, for the most part, just breeders sending out DNA to different labs to get different coat color results or different known eye disorders or like that done. But it was one and done--

Host: Sure. Mainly big disease processes that could affect the line kind of thing.

Adam Boyko, MS, PhD: Right. And there was one mixed breed dog test that was out there, but it only had-- well, I guess there were two-- but they only had dozens or at most like 300 markers across the genome. And so, in my lab, we could see-- because I had a mixed breed dog at home. And so, I inserted that dog into the test. And I could pick out which breeds were actually in that dog when I looked across the 170,000 markers that we were analyzing on the research grade array that we used in my lab and other research labs.

Host: A hundred seventy thousand. I heard that correctly.

Adam Boyko, MS, PhD: Yes.

Host: Wow. Okay.

Adam Boyko, MS, PhD: And so, I was like, if we could actually get people to test their dogs on this research grade platform, then we could give much better results about their dog. They could get all of the results for all the different coat colors and known genetic defects, and also get really accurate results for which breeds are in their dog. And we will have the genetic data now that could potentially be applied to future research studies. And a lot of people were really interested in having their dog take part in research, and they fill out surveys, they upload pictures of their dog.

And so, when we got into the incubator here at Cornell, we had angel funding and seed funding. And within the first year, it was about 11 months into the company, we actually finally launched the test commercially. And there was lots of interest.

Host: That seems pretty quick.

Adam Boyko, MS, PhD: Yeah, it was. We had a really good bioinformatics programmer that was working full-time for the company. My brother and I co-founded the company, so that was kind of fun. I got to keep working with my brother. He decided he didn't want to go into academia after he graduated.

Host: That guy again? Yeah. Nice.

Adam Boyko, MS, PhD: So, he went from being a grad student-- so basically, nobody listened to him-- to being CEO of a company.

Host: Nice.

Adam Boyko, MS, PhD: That was a big change.

Host: It's a step up.

Adam Boyko, MS, PhD: Yeah, big change for him. And after a year of doing testing, we realized that dog DNA tests were a seasonal business. Most people bought a dog DNA test for Christmas, either for themselves or for someone else.

Host: Oh, sure. I like that idea.

Adam Boyko, MS, PhD: So, we had a really big thing there. And also, lots of people like to tell us about their dogs. So, we started then taking the answers from the questionnaires that the people had given us and run mapping studies based on them. So for dogs that the owner said, "Dogs did this versus didn't"--

Host: Like personality traits kind of thing?

Adam Boyko, MS, PhD: Tried it with personality traits. We didn't see any strong signal at first with the personality traits. But we did see a strong signal at first-- and again, this wasn't why we founded the company. I had no idea that this was going to be what we saw-- for blue eyes. So, dogs that have blue eyes shared a region of chromosome 18, which hadn't been associated with pigmentation before. And we found that there's a 99,000 base pair duplication in that region. So, the dogs have like an extra copy of that part of the genome there. And if they have that, then they're highly likely to have blue eyes.

Host: Blue eyes.

Adam Boyko, MS, PhD: Or heterochromia.

Host: Wow. So, two different colored eyes or sometimes different colors in an eye.

Adam Boyko, MS, PhD: Yes. different colors in an eye or different colors between the eyes. And so, this is the blue eyes in Siberian Huskies in particular. So, a lot of sled dog breeders had been using the test. And so, we had a lot of sled dogs in the database. And so, that kind of drove this association signal.

So initially, we had about 2000 dogs with different eye colors that we could use, and we found that, but then we could reach out to owners who hadn't told us about their dog's eye color and be like, "Hey, we think your dog would be interesting for this study. Can you please tell us which one of these eye colors most closely match your dog?" And sure enough, we got a response from another 3000 individuals. We could confirm that the association was where it was. We can confirm that the duplication is what's driving the association. And we published the research. Like, we didn't want to like port it. And so, it was the first ever genetic discovery made in direct-to-consumer testing in a non-human species. So, it was kind of--

Host: That is so cool. So, the research continued because you had this product and all these people-- I know I want to talk about my pets-- shared all this information. You kind of pooled all this amazing data. And through pooling that data, you discovered this duplication and it led to this amazing discovery.

Adam Boyko, MS, PhD: Right. And there was no federal funding that went into it, right? This is all people paying for their dog's DNA tests and telling us about their dogs. And that powered the research. We didn't have to ask the government, "Hey, can I do a study on eye color in dogs?" And then, justifying that research.

Host: Yeah. I love this story, because I love that people also wanted to share their data. And then, they really, really contribute to the science that we will now share with the world. I love stories like that incidentally. I'm also going off script again. I DNA tested both of my dogs. Both I'd like to refer to as shelter specials. And I got some very fun results. One was very interesting. I'll have to share them in our little like after-podcast info.

But okay, folks, I've been waiting, like, weeks to ask him about this. I went back to your site because I received the results on my latest rescue, who I adopted just a couple of months ago, and found out that you can do an age test. Now, my age test is pending for this new dog. Because I really, really want to know, I'm stalking the website. But can you tell me about this? Because this blew my mind. I didn't even know. I've been a veterinarian for almost 19 years, and then went to your site and looked at this and was in shock and ordered it immediately.

Adam Boyko, MS, PhD: Yeah. So, it's a new test. We only launched it a couple years ago. And the reason why we did it is, when you're offering a consumer product, you really want to listen to the customer and you want to know, like, "Do you have the right product for your customer?" And we have 4.6, 4.7 stars on Amazon. People love the test. But we still wanted to reach out to customers that had complaints and be like, "What would you have wanted more out of this test? Why did you not give it a five stars?"

Host: By the way, the test is very easy to do. I actually really appreciate, just like swabbing your dog's cheek. It was really great.

Adam Boyko, MS, PhD: Right. right. Some dogs handle the swab easier than others. But most dogs, if you just wave a treat in front of their mouth, they start salivating and you can get the sample from that. And of course, my dog at home was very well-trained because every single time we upgraded the system, we used her as a practice dog. So like as soon as she saw the swab, she would sit up and she knew she was getting a treat at the end. So, she was great with it.

Host: Pavlov moment. Yeah.

Adam Boyko, MS, PhD: Right. So, the three main reasons why customers had complaints about the test was, first of all, they would say, "You got the breeds in my dog wrong." And we would go back and look. And 99% of the time, that was not the case. There was a bet in the family and they lost the bet. If they win the bet, they are five stars, like over the moon, everybody else was wrong, I knew I was right.

Host: That's so funny.

Adam Boyko, MS, PhD: So anyway, that's the first class. We weren't having a problem with the algorithm. The algorithm was working fine, which people just wouldn't believe it.

Host: Got it.

Adam Boyko, MS, PhD: The second one was I thought the test would tell me what my dog is allergic to. That's not really written in the DNA. That's a complex sort of situation with the immune system and what it's been exposed to.

Host: Agreed. People, everywhere talk to your veterinarian about this. There's actually not a simple test. There's different tests for different things. So, yes, got it.

Adam Boyko, MS, PhD: Right, right, right. And I mean, some people are more predisposed to allergies than others. And some dogs are going to be more predisposed to allergies, but that's not going to tell us exactly which allergies are problematic for that dog.

And then, the third one was, "I thought you would tell me how old my dog was." And that's a trickier one because the test we launched with looks at the dog's DNA sequence. And your DNA sequence doesn't change over your life. It doesn't matter if you're testing your dog at nine weeks or nine months or nine years, the result is going to be the same.

Host: So when you say sequence, you mean the order of the DNA, it's all the same always.

Adam Boyko, MS, PhD: Right. Just the letters A, C, T, and G. Dogs have two and a half billion base pairs of DNA per genome. And so, we're looking at the ones that are driving different traits and ones that are informative for ancestry and all of that kind of stuff. And those aren't going to change. However, DNA is not just those base pair letters. There's modifications to DNA, which we call epigenetics, right? So, the DNA is the same in a blood cell, in the white blood cells in your saliva, in a muscle cell, in a skin cell. This DNA sequence is the same. That's why it doesn't matter what kind of sample we look at, the same dog is going to yield the same DNA profile.

Host: Got it.

Adam Boyko, MS, PhD: But of course, those cells are all very different and they're very different because some of the genes are turned on in some cells and some of the genes are turned off in some cells. And of course, in an infant, the genes that are turned on and off are different than in a middle-aged individual or in an older individual. And so, there are regions of the genome that are regularly changed throughout the life of an individual either becoming more methylated or less methylated. And so, this is a modification that doesn't change the DNA sequence, but it changes the ability of the genes within that sequence to express themselves.

And so, we could design a platform that looks for those types of modifications, and then build a model to infer what the age of the dog is. And so, in this case, if you test a dog at nine weeks, nine months, or nine years, it's going to give a very different result because even though the DNA sequence is the same, the amount of methylation in the different regions is going to be different over time.

Host: Was that well-known? Like, did we all know about methylation or was that something you discovered in this process?

Adam Boyko, MS, PhD: Well, no, we didn't discover it. We fine tuned it for dogs. We look for dog-specific regions that would be informative. We're really interested in knowing why some dogs age faster than other dogs and whether there's environmental factors that accelerate this. But what people are most interested in is knowing approximately how old is my dog.

Host: Yeah.

Adam Boyko, MS, PhD: And so, the test that they wanted, we knew we had the science that we could build it. We went out, we built a very large reference panel of dogs of different ages and different sizes with known birth dates that we had samples from because they had already sent us samples for doing the Embark dog DNA test.

Host: Using the information you had.

Adam Boyko, MS, PhD: Yep.

Host: That's very cool.

Adam Boyko, MS, PhD: So, Cornell has a biobank of samples, so we could look at that too. We could make sure that our inference worked regardless of whether we were doing it from a saliva sample or a blood sample or some other kind of tissue.

Host: And this is accurate up to, what, a year or 18 months?

Adam Boyko, MS, PhD: Yeah. So, most dogs are getting a result within a year of their actual--

Host: Birthdate, right?

Adam Boyko, MS, PhD: Actual birthdate. And that's when we talked to veterinarians and they said, that would be really useful because people are adopting dogs and they'd have no idea how old that dog is. And if I know plus or minus a year how old that dog is, I can figure out how to treat the dog and what to look for better.

Host: Yep. It really is so helpful as a practitioner, I have to tell you. And that's part of the reason why I decided to seek out this test and order it for this new dog as well. I found some things on his exam. And was like, "Hmm, maybe he's a little older than I thought. Maybe that means for me that I recommend blood work earlier to an owner, right? Even though they might seem okay, right? That way, we get some baseline info as they age further. So, sometimes it's really deceiving. So, I'm very excited. Two to four weeks. I just got the email two, two to four weeks, and then find out how old Bill Hodge is, my labradoodle, because you also told me he is 50% poodle and 50% lab.

Adam Boyko, MS, PhD: And you knew that when you adopted him?

Host: No, he was found roaming the streets in another state and someone found him. And they tried to find an owner. And no, I did not know. He kind of looked it, but I'm not a betting person like some of the people that, you know, you've heard some comments No

Adam Boyko, MS, PhD: I've wrong by guessing based on a dog's looks.

Host: Right.

Adam Boyko, MS, PhD: You have to have the DNA to tell you what breeds are in the dog.

Host: Yes. I'm also happy about this test because, as a practitioner, you know, we often get the question like, what do you think the dog is? And we just kind of go by exactly what you were saying, like coat color, maybe like kind of how curly the coat is or how tall they are. If they're brindle, we might. So, this is very, very helpful for us. And so, have you discovered in all of this-- because I do really think that all of this is amazing. Have there been any findings that you really think have helped, like canine health, just in general, like is there something you discovered that you're like, "This is great because people get answers, but look at what we've discovered for dog health."

Adam Boyko, MS, PhD: So, I think that the two ones that come to mind first. So, the blood work for the dogs. So, clinicians order blood work frequently. They try to use it to rule in and out different possible diagnoses and figure out whether the dog needs treatment or what to treat the dog with. In my lab, prior to starting the company, we used the biobank dogs to kind of see whether any of the blood values had genetics underlying them that might be of interest. Because we know amylase, which is measured in these blood chemistry panels was selected for as dogs adopted a modern human diet.

Host: Oh. Diets, yeah. That makes sense.

Adam Boyko, MS, PhD: So, humans and dogs both have increasing copies of the amylase gene so that they digest starch more readily, because now we're having a much more starch-heavy diet and a modern diet than we had when we were hunter-gatherers, right? And so, we saw that effect. But we also saw a really strong effect with alanine aminotransferase, which was not expected, ALT.

Host: ALT.

Adam Boyko, MS, PhD: This is also a value that comes from the blood chemistry panels.

Host: And in practice, we look at this a lot for liver indications for those of you listening.

Adam Boyko, MS, PhD: So if you have an elevated value, that's an indication that there's possible liver damage disease and treatment might be required for that. But we saw that there's a subset of dogs that have a mutation on chromosome 13 that seems to give them baseline low ALT.

Host: Wow.

Adam Boyko, MS, PhD: So when these dogs have the liver problem, their ALT does elevate, but it typically doesn't elevate outside the normal range.

Host: Oh, just blew my mind, because like for us, low ALT has no real clinical significance, but it has huge clinical significance if they have low and then they go up, but they're still in the normal range.

Adam Boyko, MS, PhD: Right.

Host: That's incredible.

Adam Boyko, MS, PhD: So now, we can tell owners when they take the Embark test whether their dog is one of those dogs. And so, they can slide that with their veterinarian. Maybe do baseline blood work initially to get what's normal for their dog or, at the very least know, well, let's look at other indications of liver damage because we don't really trust the ALT value elevating.

Host: As a clinician, that really, like-- again, he made my brain explode, everyone-- but that's such valuable information because if it's low normal, then at least we can even trend it. And then, if it's going up and then we keep an eye on the other, like, liver function tests and other values that represent the liver.

And I was just thinking about, if you discover, let's say a breed has Doberman in it. And we might consider testing because there's a very well-known disease in Dobermans with bleeding, something called von Willebrand’s that maybe you might have people that say, "Hey, let's maybe test for an issue with von Willebrand’s factor," if we now know this dog has a lot of Doberman in it.

Adam Boyko, MS, PhD: Right.

Host: Is that something also we might consider?

Adam Boyko, MS, PhD: No, absolutely. And we screen for the known genetic defects that caused various von Willebrand’s disorders for the different breeds. So, you don't have to know which breeds are in your dog to begin with, for us to do that genetic screening. And of course, clinician might want to followup with a clotting test just to confirm that the blood seems to have that. And that has saved countless Embark owner's dog's lives. I mean, we've gotten emails, "Oh my God, thank you for telling me my dog has a bleeding disorder. We were just about to get it spayed. Now, we're going to like hold back." Or in another case, about six months later, the dog got hit by a car. And the owner knew not to go to the regular vet because they didn't have transfusion there, and went to the larger veterinary practice that did have it to make sure that their dog was going to be able to pull through. So, just knowing that you have a dog that has this genetic tendency can be extremely valuable.

Host: I want to make sure everybody listening understands this, because I think is a huge impact on how we can treat animals. So yes, we think it's adorable. We want to find out like the dog breeds. I'm talking about age. Yay. It's cute. You want to see if you could win a bet or see like what kind of traits your dog has. I love hearing this. But we have these really important genetic factors that impact health, like bleeding. And we know we do surgery, we spay and neuter. You were saying they could have trauma. And this could mean life or death for an animal if they have a factor and we know that they're positive for this, that like you said, that client can go to an emergency hospital and say, "Hey, I'm worried my dog will need a transfusion." So, they kind of don't take the time out. We don't want time wasted when it comes to emergency medicine. So, I love this on all levels from general practice to our emergency situations. That's really wonderful to hear. Wow.

Adam Boyko, MS, PhD: And then, the other one I was going to point out was we had a lot of Rhodesian ridgeback owners that we're testing. So, they have this inherited deafness, early adult onset deafness that's recessive. And they were trying to get rid of it in their breed, but they didn't know what the genetic defect was. And so, we could look at our database and they contributed some dogs that were deaf to help us identify it. And we could identify a region of chromosome 18 that was associated with deafness within the ridgebacks. And we could compare that to our much larger data set of dogs from all different breeds. And we could see that it was actually, for some reason, that stretch of DNA in the deaf Rhodesian ridgebacks was most closely related to a stretch of DNA that some shihtzu had, which was sort of odd. But of course, the shihtzus aren't deaf. And so, we could sequence across the Rhodesian ridgeback sequence and the shihtzu sequence and identify the genetic difference between the two that is leading to the deafness. So by having a really big database and by having samples that were contributed from known deaf dogs, we could go through, we could identify the defect that causes the deafness. Now, it's part of the test, so all the Rhodesian ridgeback owners can test for it, and they know that if they have a dog that's a carrier, they know not to breed it to another carrier. And now, we don't have deaf Rhodesian ridgebacks that are being born anymore, which is great because a Rhodesian ridgeback is a working dog. And so, we train the dog, then it gets into adulthood and it becomes deaf and it can't work. And it just is really frustrating for the dog as well as the owner.

Host: Yeah. And this is part of responsible breeding. We hope that people are doing-- like, they're not only looking for-- a lot of us, as I grew up as a veterinarian, we would test for some big things in breeding. I'm going to say big things, right? This is a very big thing. But let's say we looked at their hips, we looked at their eyes, we had to check their hearts. But some of these, I couldn't look at a change in a chromosome that would lead to deafness later in life. So, breeding this out so that way people who have these dogs that they use to work are at an advantage, right? Because we don't want that in the line. So, that's really cool.

And so, now, we've had all these impacts on canine health. Do we have any impacts? Like, have you seen this kind of translate to the human side? Because I know you talked-- at least you mentioned this in your bio, why are dogs such a good model for some diseases in humans and like in human health? What can we do with this information?

Adam Boyko, MS, PhD: Yeah. So, a lot of model organism studies for human health are done using mice, which you can have mouse colonies and you can do lots of experiments in mice. And they're good for many things. But mice are small. Dogs are closer in size to us than mice are. Dogs have a more similar diet to us. They live longer than mice. They don't live as long as people, but you get sort of the same aging sorts of effects that go on. Genetically, dogs are more similar to humans than mice are.

So if you have a genetic defect in a dog, it's more likely it's going to map to a region of the human genome, and it's more likely the effect of that genetic defect is going to be similar to what that defect in humans would cause than in the rodent model. And of course, the dogs are spontaneously developing cancers and heart disease, whereas in the rodent models, you usually have to inject it, right? And so, you have this spontaneous model species that's being diagnosed with the same disorders that humans are being diagnosed with in veterinary clinics that are using many of the same diagnostics as human, you know, medical facilities. And so, there's this real great opportunity. Some of the diseases that are rare in people are common in dogs. And so, it's a lot easier to study that in dogs. And then, we can apply it to humans. In many cases, because dogs, a dog breed doesn't have as much genetic diversity, it is easier to find the defect that's in it. So, this disease is common in this breed. And now, we can find what the gene that's broken in the dogs that get the disease in this breed are. And, "Oh, isn't that interesting? We didn't realize this gene was involved in this disease process." Now, when we look at the human data, we can see, "Oh yeah, 20% of the time when we don't know what the genetic causes for this disease, it's actually a defect in that gene." and so, this has happened time and again for everything from narcolepsy to skin disorders. We found it first in dogs. And then, we knew to look in people, which is kind of fun.

And then, what's nice is when you've got this really big dataset of dogs that have been tested for all of the known inherited disorders. So if you have a dog that's, "Hey, genetically your dog is going to get this disease," if that dog goes on to not develop that disease, there might be something in that dog's genome or in that dog's environment that we want to learn about for how do we protect from this. And so, that's another area of research where dogs can really help crystallize hypotheses around therapeutic targets for different human diseases.

Host: Yeah. Gotcha. I want everybody listening to understand that it's not easy to study diseases. And we use mice and rodents because we have a fair amount of them. They reproduce at a certain rate. You can pass on these traits and study them fairly quickly. But the dog being similar obviously is really important for our learning, not only for the canines, but for humans, but not as easy. I'm not commenting on studying disease in dogs in a lab. I am just commenting that what a neat way for you to take an opportunity to study genetics and use it to help in human health. That's my comment, everyone. I think that sometimes, when people hear about studying animal diseases to help human health, it has this really negative connotation, but it's not always like that. And this is a good example of that.

Adam Boyko, MS, PhD: Right. All of this research is done with pet dogs. They're getting diagnosed in clinics.

Host: You're not inducing anything. You're not giving them any disease. They have them already. When he said spontaneous, it means they got the disease on their own and now we're using this to our advantage.

Adam Boyko, MS, PhD: Right. Otherwise, it's just wasted. The scientists--

Host: Wasted information. Yes.

Adam Boyko, MS, PhD: So, it is just enabling research that wouldn't otherwise be possible to get that.

Host: Yeah. And the other thing that I think is great that you mentioned is, yes, we're studying the animals with disease, but with those that have markers-- and he just said-- where we're like, "Uh-oh, they're going to get this disease and they don't get it. That's also incredibly valuable information." Like why? Science is always providing more questions, and I'm not sure how I feel about that, but here we are. It's why we do what we do.

So kind of changing topics, you have this course that I mentioned that you teach. And it's called Genomic Data Science in Dogs. So, that's a long title. Can you tell me like, first of all, what is this course? Because I believe you designed it. Who is it for and what are you teaching? Like what are we learning?

Adam Boyko, MS, PhD: So, there are lots of students, undergraduate students, pre-undergraduate students that are interested in data science. Maybe they took a statistics course.

Host: Oh, I blacked out again. Sorry. He said statistics. Okay, I'm back. I'm back.

Adam Boyko, MS, PhD: They want to learn more about data science, but they actually want to be able to do research.

Host: When you say data science for our young listeners, could you briefly tell them what that means?

Adam Boyko, MS, PhD: Sure. So, somebody gives you a data set and asks you to make sense of it, right?

Host: Just a bunch of details and they're like, "Here."

Adam Boyko, MS, PhD: Right. And it could be everything that a very simple thing where it's like, "What is the mean and standard deviation of this distribution? Is there a correlation between this variable and that variable to super complex data science questions like, if we have a trillion transactions that we've recorded by our company and customers over the last 50 years, Can you tell me when I should put toilet paper on sale?

Host: Oh, gosh.

Adam Boyko, MS, PhD: Right.

Host: Yeah, that sounds complicated.

Adam Boyko, MS, PhD: Right, right. But that's what data science gets paid a lot of money to do by some companies. And a lot of science is data science, right? We're collecting data, we're cleaning data, we're analyzing data, and we're addressing whether or not the data reject the null hypothesis or not.

Host: Sure. So, we have this idea, we're trying to answer this question. Does the data kind of answer that question or the answer we're proposing? I want to make sure we're on the same page, or it doesn't, okay?

Adam Boyko, MS, PhD: Right. And I think in high school, at least in my high school and even in my undergrad, the data science courses were very different from the others like biological science courses, right? There wasn't a lot of overlap. And so, I could learn about biological concepts and genetic concepts. And I could learn about data concepts. But I never got the opportunity to like take a genetic data set from the real-world, like a published data set that's available or the Embark data set or something like that, and then be able to ask questions about it, right? And so, we wanted to do this class where, "Okay, so you know a little bit about data science. Let's teach you a little bit more. And let's show you how to use a program in this case, R, which I think most high schoolers have access to. It's a free program. People use it for data analysis.

Host: Very cool.

Adam Boyko, MS, PhD: Let's also teach you about genomics, specifically dog genomics, so that you understand what's a gene, what's a chromosome, what's a genome, what does it mean if something's recessive or dominant, where did dogs come from, how are breeds of dogs related to each other. And now, let's show you, "Okay, here's a data set from golden retrievers. And I want you to take this data set from golden retrievers and tell us where are the genes that control the color of the golden retriever? Like, is it a cream golden retriever or a red golden retriever, or somewhere in the middle?" And so, we call that a genome-wide association study. And so, we go through the steps of how do you do a genome-wide association study? And now, you get to do that on this real data set. And by the way, we give you all sorts of other kinds of data on these golden retrievers so you can go through and design your own project and make your own discoveries with that. And then, the next week, we'll look at a different data set. Maybe it's a data set of canine cancer sequences or gene expression or something like that.

And so, what this does is obviously if you're really interested in dog genetics, it gets you really started in dog genetics, but it gives you the skillset to be able to grab a data set from the internet, load it up into a program, analyze it, make sense of it. You have some tools now that you can use. And so if you wanted to join a lab to do a research project on any species, you now have some skills that you can do that, and you'll have to learn the specifics of that species maybe to do some of the, like, how many chromosomes does a parakeet have? I don't know offhand. But dogs have 39 pairs of chromosomes. That's something that we learn in the course. And so, that helps you think about how we analyze all of the different chromosomes and where's the X chromosome and where's the Y chromosome and how do you think about a mitochondrial chromosome and all of that kind of stuff.

We go into the details with dogs, but you have the general background now where, first of all, is genomics something that's really interesting to me? And then, if it is, now can I find a lab that's doing things in genomics that are of interest to me and I can already hit the ground running because they know how to do some stuff? And that's going to make it a lot easier to contribute to the lab and grow my scientific expertise. So, just going from, "Oh, I have interest and I can work in a classroom setting" to, "Oh, I can actually work with real data sets and make real insights."

Host: That's awesome. I mean, I think about being in school and doing like little Punnett squares and being like, "This is so much fun," but this well beyond it. I like that they get to basically really fuel their passion and see if they want to go on and do other things.

This is great. People are using these programs and I forgot to turn my microphone on before this podcast started. So really, it's so similar. It's so similar, their experience with these data

Adam Boyko, MS, PhD: sets.

Yeah. So, we offered it last summer. And the response was really, really positive. And so, we're doing it again. This summer, I co-teach it with another couple professors here at the vet school. So, it's not just me talking to the students. They get to interact with several different professors and work on different data sets.

Host: That is awesome. Plus, that's networking. They meet other people that are interested in genetics. They meet you and other faculty members. So, I really like hearing that. So for our young scientists who think they're interested in genetics, do you have any advice for them? Maybe they're in junior high school, maybe they're in high school, maybe they're in undergrad, and they didn't think they'd like genetics. And then, they were like, "No, wait, I do. I would like to do more." and it doesn't have to be advice about science. Like, what do we tell these young genetic lovers?

Adam Boyko, MS, PhD: I mean, there's so many different areas of genetics. Like if you come from my background where I like building computer programs and solving math problems, or you come from the background of, "I like being able to do reactions and precipitate DNA and do complex imaging" and things like that. So, you have wet lab and dry lab scientists. They all contribute to it. But the thing they have in common is they love to solve puzzles and they love to think analytically about what is the experiment I can do that's going to tell me whether or not this is true, right? How am I going to test for this? What do I need to control for? Because we don't want to do experiments just for the sake of doing experiments. We want to do the experiment that's going to tell us is it A or is it B or is it something we hadn't even thought of yet? Like, we need to be able to rule in and out different hypotheses for science to advance. We can't just go through the motions. So, it's this puzzle solving and thinking strategically and analytically that is so crucial.

Host: You just said it, there are lots of aspects, like people can be interested in genetics and they don't necessarily have to do what you do. They could be doing other things. And again, it's another shout out to science, because that's what I love about this. Because what you do in science, it's going to be very unique to you, it might not be like everybody else. And so, that's really neat.

Okay, we've soaked in the science. And now, I'd love to just have a little little round of a couple of questions so people just get to know you before we sign off. So, I would love to know, do you have a hobby that not many people know about? Because scientists are people too. I keep trying to convince others about this. We do have other interests. You're not at dinner every day talking about genetics. You do other things. So, is there a hobby that we would love to hear about?

Adam Boyko, MS, PhD: So, I got my pilot's license in 2019, so I fly around a little bit, which is a lot of fun. It's a different way of thinking than genetics.

Host: I will say that's wild.

Adam Boyko, MS, PhD: Yeah.

Host: That's very cool.

Adam Boyko, MS, PhD: I thought that teaching genetics to vet students was difficult because a lot of vet students, they don't like the data science part. And there a little bit of data science that goes on with genetics. And I was a little frustrated. And then, I started learning how to fly and I realized my instructor sitting next to me could die if I did something wrong. And that's actually a lot harder teaching. And so, the way they think about teaching in the FAA and certified flight instructors, it really adds to how you can think about teaching in other contexts as well.

Host: Absolutely. As a person who taught live surgery, I feel like there were some components there that are very similar because you're like we have to address any bleeding and things like that. But actually, a book I recommend to students is called the Checklist Manifesto. And Atul Gawande actually studied like pilots and how they go through these lists when they're flying. And this is a read I recommend for anybody. So, I really like hearing that. Oh, okay. Wow. Well, I can't top that as far as my hobbies. But in case anybody wants to know, I am trying to learn how to crochet. Little safer.

So, I asked this last episode and I'm just curious, if I was forming a zombie apocalypse team, what trait do you bring to the team? Tell the listeners they need to know. Well, flying is pretty good, but aside from that.

Adam Boyko, MS, PhD: Flying is nice. I'm pretty good at gardening.

Host: Growing your own food.

Adam Boyko, MS, PhD: My garlic bed does well every single year. So if the zombies don't like garlic, I don't know-- that's vampires, I guess.

Host: It's true. And then, when you are in the lab, when you're doing your work, no one else is around you, is there music on yes or no or anything on in the background?

Adam Boyko, MS, PhD: It depends what I'm doing. If I'm just coding, yes, it's generally classic rock. But my kids are getting older now and I'm branching out a little bit more, so, you know...

Host: Nice.

Adam Boyko, MS, PhD: Yeah. My middle son is in the top 0.05% of Bad Bunny listeners, so that's going on. I think I'm ready for the Super Bowl. We'll see.

Host: Yes. Love it. That's amazing.

Adam Boyko, MS, PhD: But if I'm writing a manuscript or something, no, I can't really do that with the music. It gets too distracting.

Host: Yeah, I totally agree with that. I tend to have silence if I'm really, really concentrating. And then, sometimes, other things in the background. Okay. And last thing, Star Wars. I'm back to this, Dr. Drew Miller, if you're listening to this podcast, Star Wars or Star Trek?

Adam Boyko, MS, PhD: Definitely Star Wars.

Host: Whoa. That's the right answer. We started 2026 off right. That's all I needed, Dr. Boyko.

Adam Boyko, MS, PhD: I love all nine movies equally.

Host: Valid.

Adam Boyko, MS, PhD: Yes, Star Wars.

Host: That's a valid comment. I hope you did like the Andor series. Because as a Star Wars nerd, I did really enjoy that. Putting it out there. Dr. Boyko, thank you so much for joining us on the Cornell Vet Podcast today. Thank you everyone listening. I hope you had a peaceful New Year's Eve to New Year's Day. And I hope you set some great intentions for the upcoming year. And one of those intentions should be to listen to this podcast and learn about all the cool things we do at the Cornell College of Veterinary Medicine. So, like and listen on your favorite platform, and we will talk to you soon.

Adam Boyko, MS, PhD: Happy 2026!

Host: Yay.