Kelli Retallick is President of Angus Genetics, Inc, a subsidiary of the American Angus Association. AGI was established in 2007 to provide services to the beef industry that assist in the genetic evaluation of economically important traits.
In this episode, we start with the basics of Expected Progeny Differences. We then discuss the continued need for phenotypes and measurements, even as genetic evaluations incorporate more genomic information into their calculations and algorithms. We talk about new EPDs coming down the pike for better reproductive efficiency selection tools. We touch on gene editing technology and even get philosophical on whether all this science diminishes the "art" that many still believe should exist in cattle breeding.
This episode may best resonate with those in the seedstock sector, but it's a conversation that should be good for other industry segments, as well, as we strive for the most practical, profitable genetics possible in the beef industry.
Angus Genetics Inc.
Kelli Retallick is President of Angus Genetics, Inc, a subsidiary of the American Angus Association. AGI was established in 2007 to provide services to the beef industry that assist in the genetic evaluation of economically important traits.
In this episode, we start with the basics of Expected Progeny Differences. We then discuss the continued need for phenotypes and measurements, even as genetic evaluations incorporate more genomic information into their calculations and algorithms. We talk about new EPDs coming down the pike for better reproductive efficiency selection tools. We touch on gene editing technology and even get philosophical on whether all this science diminishes the "art" that many still believe should exist in cattle breeding.
This episode may best resonate with those in the seedstock sector, but it's a conversation that should be good for other industry segments, as well, as we strive for the most practical, profitable genetics possible in the beef industry.
Angus Genetics Inc.
Well, hello and welcome to episode number 17 of practically ranching. This week's guest is Kelli Retallick. Kelli is president of Angus genetics, Incorporated. And a fair warning to everyone who may not necessarily be directly involved in the Angus seed stock production side of our industry: this one may get a little techie. It may get a little deep into the weeds, but, I think that it's a good one to have. I think it's a good one for all of us to get to participate in and listen to. And yes, we get into the science, we get into statistics, we get into math and, and a lot of these methods that are associated with the, calculation of our EPDs and dollar values in the Angus breed and really in, all breeds. We also get into some questions like: is cattle breeding art or is it science? And, a lot of the philosophies that go into not just the adoption of, and the use of, but even the, the creation of, and the calculation of EPDs and selection criteria here in, in, cattle breeding. So again, it's an important discussion. None of us necessarily agree on the perfect set of EPDs. None of us may even agree on the perfect animal, but I think all of us agree that we need to use every tool available, whether it be our naked eye or EPDs, to find cattle that are the most practical and, and the most profitable for all of us in the industry. So. Uh this discussion was a fun one for me and i hope it's a fun one for you to listen to as well. well, Kelli thank you for joining us today on practically ranching. Uh, we appreciate you time to visit with us a little bit from St. Joseph, Missouri and the American Angus association headquarters about expected progeny differences and genetic selection. we get too deep into the weeds here, give us a little background of your professional training and education and, what you've done leading up to this position and then what you what'd you do for the Angus association and AGI today?
Kelli:Sure. Well, thanks Matt, first of all, for having me on here, I feel like. Kind of on one of the most popular podcasts around every time I'm in an industry event and we're talking about what podcasts we listen to, your seems to always come up. So we appreciate it. We appreciate it. Well, I'm Kelli Retallick. I am the president of Angus genetics incorporated, which for those of you don't know is a subsidiary of the American Angus association. But to be honest, I'm a bit of a product of the, of the Angus system. To be real honest with you. I grew up kind of on a farming operation there in Southwest Wisconsin. Generational farm. And when my mom and dad kind of took that farm over, they kind of had a hay decision to make. They actually started out showing a bunch of show steers and selling a bunch of Maine and Chi's um, in the eighties. And then when they had my sisters and I come around, they kind of looked at and said, what's, what's going to allow them to raise their family and the way they want to. And it really was that Angus cow. And so I showed heifers out of my dad's replacement, heifer pen growing up, but I am a little bit of a closet steer addict. Um, have a little bit of
Matt:steps admitting we have a problem, Kelli.
Kelli:You know, have a little bit of that show, steer enthusiasm in me. But a majority, I grew up through the Angus programs like you Matt, I was on the national junior Angus board of directors and, and I always bring that up. Cause I think that was probably a pivotal moment. Um, As to why I got where I am today. It actually was probably that part on my resume right after I got out of college, why I landed at Kansas state university in their animal breeding and genetics program. when I was interviewing there and then eventually got to come on in, into that program, they said, you know, the one thing that really stood out on paper on you was the fact that, you know, you, you maintain good grades and you did well through school, but you also were really involved in the industry. And so I owe a lot of credit to being on that national junior Angus board of directors. And at that time I just thought I was having fun, but I really think it probably propelled me. Um, into the career I had the day specifically because Dan Moser hired me at Kansas state when I came on as a grad student. And then he was the first one to hire me here at Angus. And so, I did have a short stint at the American Galvian association in between there for about a year and a half, right out of my master's program at K state, and then was able to kind of what I like to think about come home to Angus here about six years ago and started in primarily kind of a member outreach, external facing role, uh, traveled all over the country, got to make great people like yourself and, and others in the industry. And really kind of start to have a well-rounded knowledge of all the different breeding systems around the country, and then have kind of worked my way up as things have changed here in AGI. And so now I just am fortunate. I get to lead a great team of research geneticists and customer service individuals to try to create genetic solutions and put the right tools in the hands of, of cattlemen to breed better cattle each and every.
Matt:let's step back a couple of years ago, as you were doing that outreach with, with Angus and with Angus genetic thing, um, what would be, you had to say two or three of the top or myths about just performance programs and EPD use in general, what, what did you hear out there sometime that just made you cringe just as an animal breeder and geneticist?
Kelli:that's a great question. I think the one thing that, that probably always hit me the hardest is EPDs got so ingrained into our language that I think we forgot what the acronym stood for once in awhile. Um, it's expected progeny difference and trying to look at and predict which animals are going to be the best parents for the next generation. I think around the country, when you get into some of those conversations, sometimes we, we change expected to exact. And I like to talk about setting realistic expectations, um, whether that's for my teammates or for, uh, people who are using our systems and really trying to stress that, you know, we have to set realistic expectations on what these tools can do and the fact that they're ingrained in population genetics. Um, we're never going to be exact, there's a reason why we don't put a one on our EPD accuracy calculation, because we're never going to be a hundred percent accurate as to where those, those expected progeny differences are going to land. But I think having the ability to have those conversations and set those realistic expectations with people, so that way they can better utilize the tools, make their selection decisions is probably one of the things that really stood out. And then from there, I think, you think about, what's the value of. You know, in this genomic era, AGI and Angus has been using genomics since 2010, and many of the other breeds around the country have been doing the same thing, um, for well over a decade now and instill, we kind of fight or have dialogue about, what's the value of phenotypes to genotypes and seems like we always get sometimes one of two camps, right? Somebody is in love with the genomics because it's, you know, they like to tell me it's the blood and you can't mess with the blood, or somebody is in love with the performance data because that's the actual measurements. But in reality, these two things work hand in glove, right, Matt, I mean, they work together in order to give us good genetic prediction tools. And so as you, as you go around the country, I think those are the things that I would say generally, as you go to all those different environments, those are probably the repeated questions or, or the repeated conversations that you would get into with breeders.
Matt:yeah, those, those are two big ones. And the first one about breaking down EPD into the, the actual words that the acronym stands for expected progeny difference. I had to chuckle it never had hit me until this summer. And my daughter, Ava showed a registered Angus steer at the national junior Angus show carcass contest, and we get this paper and we don't market any steers that aren't Coles that we take to the feed yard or the sale barn or whatever the case may be. So see or think about registration papers on a year and a half old steer.
Kelli:Yup.
Matt:He had a complete set of EPDs and I turned to whoever it was sitting beside me as we checked this thing in and read his tattoo. And I said, does it stick? have expected progeny differences and they looked at me and they shrugged their shoulders. Like that second, that second word is pretty crucial in this situation. But, uh, yeah, we, we, a lot of times just link that EPD, that set of EPDs to the animal. And we're not trying to describe the animal. We're trying to describe what it is that he or she are going to produce in their progeny and, and, that, that makes it, yeah, that there's a lot of wiggle room. There's a lot of environments. There's a lot of, you know, the dam side, if you're talking about sires, what was the damn side offering, in those progeny as well? So you, you mentioned, genomics and that's something that I think I really want to spend some time on for the sake of folks that have heard the term of genomics, but maybe aren't in the seed stock business... don't fully understand, give us a quick genomic enhanced EPD talk and what it is that genomics do for these EPDs on, on seed stock that have been blood tested and run through the system.
Kelli:Yeah. So genomics really is, it's basically just another tool, another piece of information that we can fit inside of that EPD. So currently when we think of genomics and how we're utilizing that to predict what, say an expected progeny difference or an EPD on a yearling bull that someone is going to purchase this fall or next spring, really what we're trying to do is I like to describe it as we're trying to make our pedigrees better. And if, if you've been around the seed stock business at first, when I started describing it like that, Matt, I kind of got some wrinkled brows, but basically all's, we're trying to describe is what are the genetic relationships with the genomic relationships between our animals? So, uh, Matt, you have kids and they're full sibs. Um, and the reality is, is that even though, they're full sibs, they're all a little bit different.
Matt:They're all a lot different Kelly
Kelli:but that's the reality for a lot of people. Right? And the, and the reality of it is, is, you know, even though they're full sibs and they're getting 50% of their DNA from, from Matt and 50% from their mom, um, at the end of the day, they get a different half from each one of their parents. And so the same way with our, our individual cattle. So when we take that DNA test, we take that blood test and we run it and we get all these, what we would call a SNP reads, right? Your A's, C's, T's and G's,, your basic alleles to life. Everybody, all living organisms have them. We're just basically going in there and we're saying, okay, this is my string of DNA on this animal. And how does that relate them to other animals in that pedigree? So you're always going to be about 50% related to your sire and dam, which I think makes intuitive sense. You have the sperm and the egg. They come together, right? 50% of the DNAs from the sire, 50% from the dam, but in a traditional pedigree, if we think about kind of just a traditional evaluation without genomics, we would also assume that on average, you're, you're 25% equally related to all four of your grandparents and 12 and a half percent equally related to all eight of your great-grandparents. And with DNA, we know that that's not the absolute truth. So when I go and I do a DNA test and I'm able to kind of crunch, the values and run it through the algorithms, I may find that this young bull calf who has a grand sire who is an absolute rock star for something like marbling, because he's got hundreds and thousands of progeny for ultrasound, um, IMF measurements and carcass data from the plant. Well, if I find that this young bull calf is not 25, but 30% related, so has 5% more of his genetic material is from that rockstar for marbling. You know, that young sires EPD, um, for marbling is going to go up. Why? Because that's, that's the DNA, that's the genetic code he has to then randomly pass down to his next set and his next generation. And so when we think about genomics, we really can't even put it in its own bucket anymore because it's, it's basically just going in there to make that pedigree better and, and build those genomic relationships. And that's quite frankly, Matt, where we get all the power with our genomic testing is because we can leverage relationships all over, right, all over the pedigree. Obviously you have your immediate family, but you also have half sibs and full sibs and quarter sibs and, um, distant cousins and relatives that now I can get linked into. And, and we can leverage all of their performance data, all their actual, real phenotypes, weights and scrotal circumference measurements in, in heifer pregnancy data and ultrasound measurements to then come up with a more accurate EPD. And so when we think about genomics while, um, there's a lot of steps that happen once, once you pull that tail hair out of the tail head, or draw that blood sample, or take that little ear notch with your tissue sampling unit. Um, at the end of the day, all we're trying to do is just let's track, which genes came from, which individual animals in that individuals pedigree. And let's see how they performed in the past to help us make inferences about the future.
Matt:And so, in addition to those genomics on let's, let's just look at non-parents. So yearling and coming two year old bulls that, uh, commercial cow calf customers may be buying at sales this fall, or next spring. In addition to the genomics, we would also ideally have a pedigree estimate, which is their siren damn maternal grand sire. And on back through that three generation pedigree, basically the old adage of let's say they get exactly half siren dam. So we add and divide by two on each of those EPD traits. then we also have there before or after those genomic profiles are submitted, also have that animal's individual performance relative to his or her contemporary group, calf crop the same way, et cetera, et cetera. did they perform for birth, weaning, yearling weight, carcass ultrasounds, Kronos conference, any of the traits that we're talking about? So those would be the three main drivers of that individual's EPDs as a non-parent before they have progeny submitted the system. Correct?
Kelli:Correct. Yep. You're absolutely right. You start out with that pedigree average. The way I like to describe it is that's, that's going to be just the average of your parents and then anything you add, whether it's genomics or individual phenotypes and in their correct contemporary groups, that's just going to deviate you away from that average. So if we didn't have any other information, right, you'd have, you'd only have that average parental EPD to go off of, but as I add in their individual birth weight or their individual weaning weight, that helps me deviate those individual EPDs off of that parent average.
Matt:So since we have had genomic, technology to put into this system, for the old, the old school folks like myself, that have, have always believed in the power of phenotypes of weights and records that were and submitted for analysis through the, through the national caliber valuation through the American Angus association, AGI. I always felt like that if we had that that pedigree estimates and this animal ratio to above average for a given trait, I saw his, or her EPD move a little bit up. If it was below average, I saw his VPD move a little bit down when we threw genomics in there. Now all of a sudden there's kind of this perception of this black box that we don't get to see exactly how and why that worked. And even when we just had ratios and things like that, you still scratched your head on a few, but what is the best or does it matter of submitting the second and third step? So you have your pedigree estimate. You've got, let's say a weight and a weaning weight that you submit after you've weaned that that group of kids. Does it matter whether you submit weights on a contemporary group then genomics or genomics and then weights. And does the EPD move less if you've already submitted one before the other?
Kelli:Sure. No, that's a great question. We actually get that quite often, Matt, and that in the reality of the situation, it, it just doesn't matter. Um, prior to the current evaluation system that we're in and not to get too deep, um, for everyone, but we're, we're on a single step genomic evaluation today. And, and Matt you'll remember back before 2017, we were what we called on a two step or a multi-step genetic evaluation, where we had to kind of have this robust training population. So then we could, we could fit these genomics in his correlated traits and things, but with single-step that all went away. And so whether you send in your phenotypes before your genomics or your genotypes, get back to be able to be utilized in the system where your genotypes come in first, basically that system starts from scratch each and every week. And with every little bit of information, every performance phenotype, um, that comes into that evaluation. Basically, we can retrain that weekly on the fly. And so we're not waiting, you know, 12 or 14 months for this huge calibration to take place like we had two prior to 2017. Now we get to do that real time in the, in the really cool thing about that is we really start to see the deviation in accuracy values. When we start to talk about individuals who are really concerned and really committed to the data reporting and the use of genotypes compared to those that are only genotyping, you, you don't get as much of an accuracy boost with genomics. If you're not participating in that phenotypic collection.
Matt:So now I'm going to be the seed stock breeder who is looking at his cashflow and his profit and loss, and trying to determine where we can get the same outcome for less money. trying to contract the cheapest feed I can find and still get the job done. I'm trying to find the deal on AI, semen, all of the, everything under the sun. And I know you've heard it a million times, but what good does it do to me today? When I see very, very little movement in my individual animals EPDs, as I submit data, and I've got this rockstar, that ratio to 120 some for a trait, and I've got this puke that ratio 75 and neither EPD moves much at all because we've already submitted genomics. And maybe the genomics predicted that that's exactly what they were going to do. what cause do I have for the added labor, the added paperwork in the case of ultrasound, a technician to come and scan my cattle. Why do I keep submitting phenotypes when I'm going to get the same dang information from the drop of blood on a FTA card and genomics?
Kelli:Yeah, no, and I sympathize, right. I can understand that. I call home right after a long day after they've been out working and it's, you know, they got everything ultrasounded and fed and it's already nine o'clock at night. And now you're supposed to go in, sit down and punch your records in before you go to bed. And that is, it is frustrating, especially when it comes in and you have something that ratio is really high and you're excited about it. And then you're, you're thinking, oh, I'm just going to get a cherry on top of the cake here. And then you're, you're marbling, EPD doesn't move any more than what it already has, with genomics.. And the reality of the situation is, you know, you may be able to get away with that for a year or two, but when you really get down to it, um, I like to think of genomics in its ability to interact with phenotypic data is kind of like, uh, cash and your credit card. Um, your genomics is kind of like your credit card, right? You swipe it each time you pay for that$37 tests, um, to be able to buy that upfront accuracy, to be able to buy that kind of almost upfront movement in your. But at the end of the day, if I run my credit card for months on end and I never paid off with cash, right. My credit card no longer works. And that is the reality of the situation that we're in with genomics and the actual phenotypes. If we continue just to invest in genomics and we don't pay off that genomic accuracy with actual phenotypes, whether that's, you know, birth weights and weaning weights or heifer, pregnancy, data, foot scores, or carcass data, our genomic accuracy is going to deteriorate, right? Our genomic engine is going to run out of gas and we're not going to be able to kind of pay off our bill. And so while I know it's a little bit altruistic or kind of futuristic to think, oh, I got to keep doing this because it's the right thing to do. It's not only the right thing to do, but it's the necessary thing to do if we want to continue to go down this road of trying to capture the most genetic progress that we can. you know, Myself and Andre Garcia, who is a geneticists here with us. We, we ran just a little bit of a test if you will. And basically tried to say, okay, if, if Matt and everyone else, every one of our Angus breeders told us in 2010, after genotyping came on the market that, Hey, you know, we're no longer going to send in any ultrasound data. We're no longer going to send in those phenotypes. We're just going to rely on that drop blood, or we're just going to rely on that tail hair. Um, we actually went into your database and eliminated I believe 212,000 ultrasound records in your database. Um, from sires that were born after 2010, but they had genotype progeny. And so there was 174 sires that represented 212 ultrasound records. And basically what we did is we just deleted them from the database and active. They were never here. And we said, okay, what would our average carcass, EPD, accuracy be? If everyone would have boycotted us. Right. And just said, Hey, we're no longer going to collect ultrasound data because, you know, we ha we have this blood test, we have this hair test and we got an average of a 20% decrease in accuracy, right. When those records were removed from our database. And so what that eventually does is it not only decreases your accuracy, right? So then Matt and others, your commercial cow, calf producers, can't, you know, use those EPDs with confidence. What actually starts to do is it starts to affect your genetic trend because part of trend and part of genetic change is the degree of accuracy that we have to make. Genetic selection. And we actually saw your genetic trend for marbling EPD starts to slow and actually start to deviate immediately after we removed those records in 2000 and 2010, because Hey, our tools are less accurate. We're going to take more risks. We're going to make the wrong decisions. And so that's going to slow our genetic change. And so while I hope we never get to that point, we have been trying to demonstrate what would a world, right? Without performance data actually look like. And to be honest, um, that future isn't isn't as bright as we'd like it to be. Um, and it just reiterate this need and this necessity for people to have to go out and actually collect these phenotypes as, as see separators who are really trying to drive this genetic nucleus,
Matt:and I get that and I appreciate it. And I think that's a valuable model to do and to prove why we need these phenotypes. And that's a altruistic, you know, good of the whole type of mentality. Um, the challenge is you're serving. By a bunch of capitalists and you're surrounded by a bunch of folks who are, yes, we're in the seed stock business. And we're very, very accustomed to the feedback loop being really, really slow because we AI heifer or flush a cow, we're nine months at best, before that calf hits the ground, we're probably a year plus before we're able to market or put that animal back into our herd. And then we're another year and a half to two years before they actually make any real value to the beef industry. So it's a four to six year proposition where we're used to that. But when we see the neighbor down the road or somebody in the next state or wherever that is selling some of the highest valued seed stock, sometimes as embryos or as non-parents weened, heifer, calves, whatever the case may be. are being valued in the five to six figure range with not a single phenotype submitted or even collected on them strictly because of their EPD and genomic profiles. Um, you know, I get what you're saying and I agree with what you're saying. And I'm hopeful that all of us as true breeders of seedstock will recognize that without phenotypes, genotypes are, are archaic within a matter of years, or at least less effective, but here and now is the tough part. As we see the values that are being placed, not because the Angus association has told us to, and I'm, I'm not one of these that says, well, it's you all that are to blame for this it's us as breeders. It's us for whatever you want to call it, greed or impatience or whatever the case may be, but valuing these things based solely off of. Uh, pedigree estimate and their genomic profiles that have affected that EPD. Uh, so I don't know that there's a question in there even sometimes I, I disguise a question and then I won't even claim to have done that. It's just a statement that, that it's something we have to watch as breeders and as an association, but especially as breeders using the tools
Kelli:Yeah. Well, and I think you're absolutely right, Matt. I mean, I think that is the reality that the one thing, that you realize very quickly, even though you, you get trained in all of this, you get trained in statistics, you get trained that they're genetic selection tools, and we all know their genetic selection tools there. You know, I'm, I'm not blind to the fact that these have a lot of marketing power that these genetic tools, right. Have a lot of marketing power. And, and one of the things that, you know, I never see printed in a sale catalog hardly in it, mainly because those animals are all selling at the roughly the same age, have about the same amount of information, but I never see an accuracy value in those sale catalogs and, and probably for your commercial cattlemen, um, for your people that are coming. You know, to Dale bank saying, yes, and they're coming in and they're going to buy their herd bull from you guys this fall or next spring, or whenever it is, you know, they trust you that you've done everything in your power to try to characterize those genetics to the best of your ability. I think when we start to talk about some of these, um, high-end, uh, super high-end cattle, um, that are maybe selling for six figures or, or more than that, um, that, you know, people are buying into them and they're gonna gonna try to strategically use them to try to make rapid genetic progress as fast as we can. Um, I think those are the ones where you got to start to make sure that, Hey, how much are you protecting your investment? Yeah. When you don't know as much about them as maybe, maybe you thought you did, or, you know, do we even look at the accuracy values when we land there and those sorts of things, because that's, that's the other part of this, um, that, like I said, I never see hardly an accuracy value get printed in a sale book and I'm not advocating for that, but I, at the same time, um, when you're making these large investments, those are the types of things that, you know, start to deteriorate more quickly. Um, than, than the EPD values themselves. And what happens when you know that accuracy value gets low enough where it's not as dependable, then I think you're going to see the climate shift, right? The market shift, and go back to those cattle that maybe are a bit more well-described.
Matt:Yeah. And you're right. I mean, we need to do a better job as breeders and in educating our customers and, and the buying public of. But the value of not just the accuracy figure and that's a large part of it, also the value of, of those phenotypes and why it is important to be enrolled in maternal plus and sending in entire of that calf crop to get a real picture of how they stacked up. Because, you know, I, I don't think we have time on this podcast to talk about the predictability or the breeding precision, if you will, of these multi-generations flushed females who have never had a single natural calf, of, of, um, you know, IVF production on a six or seven month old heifer, uh, is often bred to a yearling bull that was at the top of his heap for the calf crop that year based off of his genomics and his EPD profile, but may have never had. The first piece of data submitted because he was in a contemporary group of one. So, know, that's the thing that scares me the most as we look at a technology like genomic enhanced, EPDs that relatively speaking until the last 10 years have, have seen this pretty slow methodic organic growth for the last 60 or 70 years. And development in all of these, you know, one or two traits at a time, making sure that they're validated, making sure that, that the heritability estimates are where they need to be to make a decent EPD, all these things. kind of been accustomed to that ever, ever since 2010, it's like this space race in, in genetics and genetic predictions. And I, I get why, Maybe we've got some catching up to do. We've got a lot of private money coming into to the, industry in, in genetic selection and, and genetic predictions. But, um, it is concerning to me as I see some of these top end seed stock are valued and used. And yet if you go back through that three generation pedigree very few parents in that pedigree that have got any type of data to validate what it is that we think we know about them genetically. So, uh, know, that's, it's, I think it's an issue. I think it's a real issue and I think it's something I guess I would encourage if there's folks, regardless of the breed, but if there are leaders within breed associations and genetic companies out there, I think we need to, as a group. As an industry, figure out how to reward and I am not into mandates. Um, but somehow how to reward, folks who submit phenotypic data to the system. Because if everybody quits, then you all who are in charge of making these predictions is as predictive and as accurate as you can. You're gonna have to go out and do your own R and D. You're gonna have to go out and buy data or pay to collect this data. that price tag on that is going to be. Going to be pretty steep. So I guess I would encourage to submit, continue to submit that, even though it is costly, even though there is a lot of labor involved, but I'd also encourage, you know, breed associations and genetic companies to figure out ways to make sure that that data enough value the individually PD that it's worth collecting it. And maybe that's the best way we do it.
Kelli:And I, and I, I definitely get your, get your thought process there. And I think at the end of the day, the one thing about, and even though AGI is a subsidiary of the American association, right? The American Angus association is a member organization. All AGI is, is able to do is leverage the database, that the association and its members are able to build. And so at some point in time, You know, there, there is a level of commitment, right? That that has to be there from the seed stock level. If we want to continue to, to make genetic progress in some of these traits. And I think where AGI can help is hopefully leveraging some of these data sets that maybe can't even be collected, um, on, uh, kind of in a genetic nucleus type herd or a genetic company for that manner can basically leverage some of these databases that are collect some of these databases that, you know, Dale banks and others can't go out and collect on their own, um, individual herds and there's some of those traits out there.
Matt:Yeah. Well, that, that brings me to my next question. In terms of some of these, some of these new traits, you know, we started out with weaning weights and then we had and birth and year laying and, and. You know, like I said, every year too, we sh we have added another trait to this EPD profile. there's several new traits and let's speak specifically in the Angus breed since that's what I'm most familiar with and, and obviously what you are as well. Um, w what do we see coming down the pike and let's maybe get focused on reproductive efficiency. We've got heifer pregnancy PD right now for the Angus breed. I know there's work being done on a cow stayability EPD, talk to us about what we hope coming down the line soon. And then what some challenges are because of the heritability estimates on some of these fertility and reproach rates.
Kelli:Sure. So no great question. And we've, it's no secret, we've been working on. A sustainability longevity type tool for just a little over three years now. And in some people may think about that and say, wow, that's a really long time to have to be researching that type of data, but I won't, uh, tell them how long the organization's actually been working on it. That's just
Matt:I know
Kelli:So Matt, you might've been some of that worker early on. So
Matt:I was in the room. put it that way. And that was 20 years ago.
Kelli:So, you know, these traits, these traits are messy, um, and that, and that's probably the hardest part about them. And, you know, specifically let's use the Angus example where we've never had a, a mandatory whole herd reporting system or a mandatory, uh, inventory based reporting system. And rather. Held the ground and the association has held the ground that that individual members get to participate at whatever level they want to with the association, which I think from a membership organization standpoint is a really great thing from a data collection standpoint. It, it gets a little bit harder specifically when we're talking about traits like, you know, reproductive traits in Cal function traits like stability or longevity, or even have for pregnancy. You know, there's so much noise in these data points, regardless of how well we collect them, um, which obviously lowers the heritability. And it's a little bit harder for us to make genetic change. But on the flip side, there's some of our most economically relevant traits we have. And so kind of looking at kind of what we've been doing for the last handful of years around sustainability. Like we like to call it functional longevity, and here are basically how long are those cows predicted. Stay in the herd, but not only stay in the herd, but stay in the herd and have a calf every year. I think you would agree with me, Matt, if a cow's just hanging out and she's not having a calf and she's not weaned off a calf every year. Um,
Matt:somebody else needs donor,
Kelli:she's probably, not all that successful. And so what we've been trying to do is, is leverage, um, the inventory based reporting system that we have, uh, that maternal plus program and inventory reporting through AAR that, that doctor Esther McCabe kind of looks over here now, um, and try to leverage that kind of whole herd database. But then we also know that there's a lot of periphery information outside of that whole herd database. So this periphery information that I'm talking about is just, we know we've had breeders that have been reporting individual animals into our herd book every year. problem is, is we run into these areas where all of a sudden, uh, you know, we've had a cow that's reported her first three calves, she misses two, and then she comes back and reports her reports, her six calf. And so those are the types of animals that then we have to start to make assumptions about. Um, if we want to try to utilize that data inside of inside of that, uh, inventory based reporting program, the good news is with this functional longevity and where we're at today with really any of these, these longevity stability type traits is we have more sophisticated. Sophisticated models. Um, we have better computing power, which makes things like, you know, I'm going to throw out a statistic world makes things like using random regression, um, for something like longevity, a real reality for us to be able to utilize and use that, to predict these EPDs and these genetic tools, um, on a weekly basis. And so what we're trying to do right now is basically predict, um, you know, how long that cow's going to stay in the herd and house, you know, and how successful she's been over that time. How many did you know, did she report a calf every single year, but as we really get down to it, one of the things that's going to continue to make this tool, even when it hits, um, as a research EPD, or it goes into production, The commitment from individuals to do that whole herd reporting, because quite frankly, I can look at patterns in the data, um, and I can make pretty good assumptions about it, but I'd much rather just have the farmer ranch. Tell me what happened to that cow rather than for me have to make an assumption as to what happened to this cow, that at three years of age, she dropped off and never reported another calf to us. Um, I would much rather have Matt tell me that Matt, tell me what, um, is happening to that cow than us to have to make an assumption about it. But until we get that more robust database, that's kind of in the position that we're in a little bit.
Matt:Yep. And it's shocking to me, it's disappointing to me that you even have to beg for that kind of data, because as a cattleman who. Expects the cow herd work for me, not necessarily the other way around. It's amazing. about our customers, which that's, I guess why I see that we do these types of things, why we report all this data so we can find the genetics that work and call the genetics who don't. But for this case, let's say, forget about those customers. If, if I don't record and submit data, like she didn't breed as a wet two year old for second calf. If I don't and record all of the udder scores and, and, and teat size scores. When she calves year after year, season, after season, can I ever expect to call ones that, that don't work for us and, and make my life easier as a cow manager. So to me, fertility and repro traits, yes, I understand... Low heritability estimates. Yes. I understand we can't make a lot of progress based off of lowly heritable traits, but at the same time we can make a little, and we can, I think, on the culling side, which is really what we're after, right? We're, we're not after necessarily the top 1% of the breed for We're out to make sure we don't use the bottom third or whatever the case may be. And those types of traits, I it's, it's shocking to me that that is a breed that is a set of breeders that you have to continue having these meetings and beg people to send it in. But, um, but it still appears to be the case. I think, I don't know where you're at now, as far as cows enrolled on the maternal plus whole herd reporting system or how many heifer breeding, uh, data pieces you've, you've got thus far, but I know it's not where we, where we need.
Kelli:Yeah, exactly. And I would say probably right now, if you want to, you know, look at that whole herd reporting system, we have just under a hundred thousand females enrolled, which is, which is a great number, but we have over 300,000 registrations every year. And so I know that one, there's more cows out there that are actually getting cash registered in the herd book, but there's, there's far more beyond that of the capstone. Don't actually get a registration number. And so, and the reality is, is we just want to help you, um, all characterize the genetics. You know, we get questions on heifer, pregnancy around. Why aren't we using first service versus second service versus third service somehow weighted inside of that model? Well, the reality of the situation is, is that we don't have enough, second and third and fourth services recorded in our database to actually go in there and see, okay, how maybe how could we make the model better, right. By use utilizing some of that information. And it's not that we just want more data to play with, but some of these questions that we continue to get from individuals, right. That would, you know, help you all make better decisions. Unfortunately, right. Our hands are kind of tied until we have the database to work through. So that that's one of the realities, you know, you can never put the cart before the horse here. And some of these when we start to talk about EPDs and genetics and creating tools, um, and that's really why we want the data and not to mention, I think anybody who's ever been on that maternal plus system, when I go and talk to them, um, you know, including yourself, Matt, It makes them a better. They always tell me, it makes me a better manager of my cow herd. You know, it makes me sit down at least once a year and, and go through and say, you know, who's doing their job and who's not doing their job. and quite honestly helps make those really beneficial calling decisions, that really help our Angus, producers and commercial cattlemen and anybody who's raising cows and has a cow herd just make them make and build up better cow herd for them to work off of.
Matt:Yep. And it's still the number one profit driver, regardless of what segment of the cow calf business we're in registered or commercial, those fertility and repro traits what either make us money or cost us money. Um, can talk all we want about how the kids performed and how the carcasses graded. But, uh, if we don't get mom or sister bread, none of those EPDs have any value to us. And so I applaud the association for, for doing what you're doing, and I would just encourage you to, to keep fighting the good fight on, on these reproductive EPDs because they're, they are of immense value to us as breeders of seed stock. And I think they're probably have even more value to our commercial customers that are wanting to make sure that they can, can keep those cow herds as high-performing functional, um, cows within the breed. So, or within the industry. So keep, keep it up and we'll look forward to that functional longevity PD and maybe at udder score EPDs and some things like that, even on the horizon?
Kelli:Yeah, they are. So we actually had an intern in this summer who, who worked on some utter and teat score, analysis. So, you know, you've been avid data recorder on, on udder and teat scores. And we were able to kind of take some historical data that we had through a lot of our individual aims users and other users who had maybe been recording that data, but not necessarily sending it in. Um, we launched the other scoring system, and put it on AAA log and all of our. Fields where people can actually submit data to the association here last year and saw a nice uptake in the number of records. And I would say we had about 35,000 usable records for teat and udder scores um, in that analysis we did this summer and we're able to start to work towards a prototype EPD But I think before we would get to a point where we release anything. Um, we definitely want to allow that database to girl a little bit, you know, we have, you know, several years worth of data, but there's only about 5,000 records in each year. Um, you know, kind of stemming back over the last 20 years or so Matt. And so we really want to be able to make sure, you know, we're characterizing the current genetics, the best that we can. And for that reason continue to encourage that data collection, especially with this next, you know, fall and winter calving season, coming up and hopefully bring more of those scores in so that way we can, can check her models and make sure that. We're doing a good job at predicting the current genetic base as well.
Matt:Great. Well, that's stuff. And, and, you know, you mentioned that all you're wanting to do is, is better characterize the breed for those of us that are hopefully propagate the good ones and, and select against or call the bad ones. think that's something that we as breeders, we, as people in the beef industry, um, sometimes forget and want to cast blame on someone else. And quite often, it's the breed association. The breed association pushed me to breed this type of cattle. The breed association told me that this was a good EPD and that was a bad level of EPD and the breed association devalued my bull or cow when they released Milk PD, whatever case may be. Um, makes me chuckle every time I hear that because to me every time there's a new EPD that characterizes the population of beef, cattle, genetics, as long as that EPD has been researched. And as, as long as the best effort science and scientists better characterize that, that population, then it's a free world for me to decide whether I want to place any selection, pressure on city PD. if so, how much? I look at this like a great big old smorgasbord buffet. I may not, I may not take the steamed spinach when I see there's t-bone steak right past it, or the chocolate eclair at the end or whatever else. I don't, I don't have to have every single EPD dialed into my selection matrix at the top 1%. I may say, you know what, and for my customers and in my environment, this or that EPD has absolutely no economic relevance to our management scenario. So I'm going to let it float. I don't care if he's the best or the worst, because it's not my matrix. conversely, this one, you know, if I'm breeding heifers, obviously calving ease, I'm gonna put some selection emphasis on. So I think that's one thing again, maybe I'm preaching to the choir here, but as breeders, I think we have to remember that it's not the association telling us to use the CPD or telling us that we have to put this at the top of our list. It's you characterizing beef cattle population then allowing us to select which ones work for.
Kelli:Yeah. And we're not working in a vacuum here. We, you know, we listened to the breeders, the members of the American Angus association, we get direction from your American Angus association, board of directors and your breed improvement committee as to, you know, what traits we need to look at and, and how we, you know, what's working, what's not working, um, what could be better. And at the end of the day, you know, I get, I always get the question, especially as a new EPD comes out and they'll say, well, how accurate is that EPD or, or how good is that EPD? And, and the reality of the situation is, is we wouldn't put even a research EPD out to the public or to the membership or to the commercial cowmen. If we didn't feel that we've done our due diligence on. On the research side of things, making sure that the estimates that we're putting out there, you know, have have low standard errors and in are fully vetted before we would even put that in front of someone. So even when we have a research EPD out there, I mean, from a modeling standpoint, from a workflow standpoint, we feel pretty confident and comfortable, right. That what we're putting out there is a good product. And so when it comes down to how well does the EPD describe my individual cow herd, for instance, um, anytime I get a question about that, or I get a question about, you know, a foot score, UPB it doesn't line up, um, with the foot scores or the actual foot conformation I have in my herd. The first question I ask them, Matt is. Are you participating in, in the data collection on that data point? Because if you're not participating, it's really hard for, for the system to be able to characterize those genetics correctly. And, and I completely understand of the time and the labor that it takes, um, to be able to collect those and send them in. Um, but at the end of the day, if we quote unquote, want to make our EPDs better, always going to do our best with the algorithms and the statistics to continue to find ways to make them more predictable and accurate. But at the end of the day, a lot of it revolves around, you know, how good is our data recording and how much information do we have to leverage within that system?
Matt:Yeah. Couldn't agree more. And that that's been my short answer to folks say, is this new EPD shed foot score, heifer pregnancy pick pick one of these more recently released EPDs is it, is it even worth the paper it's printed on? And my response is, well, it characterizes the sires that have got some progeny submitted pretty well. So that tells me that it will be it's up to you and me now it's up to us to the EPD. If we want to select for it. If we don't take the phenotypes, send whole contemporary groups in and let the science work Because it's okay, if it's good enough now on high accuracy type sires to see differences, then it's going to be really good when we start sending in large groups of data. And so that. That's my short answer. I don't have nearly as many letters after my name though, so I get away with,
Kelli:That's okay. I had a good friend tell me, uh, the other day we were in a conversation and he goes, well, you built us the tool, and now we just have to use it. So we can't make genetic progress. Right. If, um, even if we build tools, if, if no one uses it, it's kinda uh, if a tree falls in the forest, does it make any sound? Well, I guess if no, one's there to hear it. It doesn't. So, um, that, that was kind of an interesting comment by him.
Matt:So what about gene edited animals and specifically, is out there today? And we don't need to get deep into the policy of. Can, and can't be registered, I guess, in the American Angus association, but where do you see that going what's available today and what's going to happen in the future? I mean, maybe none of these phenotypes or genotypes even matter because we're just gonna Frankenstein our way into perfect beast and everybody's going to have use
Kelli:Yeah. Yeah. That's, that's an interesting thing to think about, right. That I think when you think of gene editing, the way I think about it is, um, it's kind of like the, the as Alison VanEenanam and would say, it's kind of like the cherry on top of the sundae, right? If we can, we can use all of our quantitative statistics to breed better cattle, and then right at the top, we can gene edit something that we can propagate those genetics quicker, or we can take care of, of something that's really detrimental to our industry. I think gene editing can, can have a pretty big place. I think when, when you think of gene editing and you think of something like for instance, BRD, um, in, in wouldn't it be nice, Matt, if we could find the gene that controls BRD. So we could just gene edit an animal and, and rid that off our population. Yeah, that would be awesome. Um, I would be more than happy, right, to get behind something like that. think the reality of the situation is it's a lot of our traits that riddle our industry, a lot of our big problems, I think they're poly genic. Right. And I think there are a series of a lot of little genes, right. Um, a lot of interactions, a lot of correlations, a lot of overlap and, and I'm not sure it's quite as blue sky as that, but heck, um, I'm on board as somebody who wants to continue to try to collect that information and try to weed it down, and see, Hey, can we fix some of these issues, in the industry, whether it be a. Um, welfare issue, a disease issue, a, you know, profitability issue that says, Hey, if we could just target the specific thing and try to take it out of the map, you don't take perse for instance, in pigs. that's a, that's a huge detriment to the industry when somebody catches it, there's a gene edit for purrs out there. And if they could just get it passed through the correct channels, that could be something that could be really huge for that type of industry. So I would never say never Matt, because again, I'm a geneticists and a statistician, so that means I have math with error, and so I I'm cool with that. but at the same time, I think, you know, if it's a good enough edit, if it's a big enough edit, um, I think it can have a pretty big impact. if we're just doing things like, you know, changing coat color or, or those sorts of things that are pretty easy to do with our simple breeding programs, right. By just crossing different lines. think those things might, might become a little less valuable when we start to talk about gene editing, but definitely from a standpoint of proof of concept, um, could be a really real thing that could help us move down the correct path,
Matt:well, we will stay tuned. And I know obviously there is, there's a lot of Distrust by the public, as we talk about gene editing. Now, am I correct that for the most part, most of the gene editing that we've seen, or the work that we've seen done in beef cattle populations has been within the species has it not, other words, taking, a piece from one breed, if you will, and inserting it to another breed to make that edit. So that probably is a little less, used the word Frankenstein probably a little little less Franken's yeah. Scary to the, the public. But, uh, but yeah, I mean, it's, it's ever evolving and, um, there'll be some work that way. I mean, you mentioned the, the poly genic nature of a lot of these traits. I think we learn a little bit more each year, or maybe we should learn a little more each year about the epigenetic nature of a lot of these traits and the fact that seeing some traits triggered in one region-- I think-- that don't get triggered in another or, or under one certain management that maybe don't under another, and it's not typical G x E interaction. It's something more. And, and that's one of those things that I think that, um, you know, genetic predictions, um, uh, the work that you all do at that places like AGI is, is invaluable as we go forward and try to to better select these catalogs.
Kelli:Yeah, no, I couldn't agree more. And I, you know, honestly, I I'd listened to your podcast not too long ago about the big data. And I think that's really where, where big data where it's kind of this big on structured data could kind of help us to kind of help us say, okay. Um, what things are happening around this calf's environment that either better aids us in the selection of the parents we want to place in those invites. Or helps us understand, you know, what resources, management practices do we need to put in front of those individual animals to make sure they reach their full genetic potential, whether that's something from an epigenetic, realm, whether that's something that once these, these calves get here, you know, how can we pair management practices to make sure we're getting the most benefit or, or just using something like, um, hair shedding or pulmonary areterial pressure EPDs to basically match those cattle to their different environments. And so I think that's where I can see, you know, big data kind of this big, messy data that basically if we could get it in the right spot to be able to understand, um, you know, how do we create solutions with that product? Um, that's where I see some of this maybe leading to in the future. And I think. I'm really excited to be in the place that I'm at, because I think, um, there's going to be no better, seat in the house to be able to watch it unfold and then watch our great cattlemen out in the U S beef industry. Be able to take these tools and run with them.
Matt:well, I hope we can. And, and you know, here we are talking about what, data can come into these EPDs and how do we do even a better job of, of working more information, not less to make robust and better genetic selection tools, yeah. There's a Cattleman's meeting that I attend that I don't hear someone quite often, uh, member of an breed association say, you know, these EPDs just aren't worth the paper. They're printed on them and throw them away and go back to select and good cattle. And good cattle to them are ones that are pleasing to the eye and have a certain look of a certain trait that they can see, but not necessarily measure. I don't think I'm the only one who sees that and hears that. And maybe there are reasons there are some foundation traits that, uh, that we haven't either measured or we haven't until recently had selection criteria that we could use download from, uh, a breed association select that way. But interesting to me as we get more and more information that we have a few more people that are saying, you know what? I don't want to use it. I'm going to just go back to breeding the cattle that I like to breed. Um, I've thought a lot about this. And I used to be in the camp. I cut my teeth, hearing Tom Perrier, sit in our living room, to customers and talking to other breeders about what contemporary groups were and 205 day adjusted weights. And you know, this is in the eighties. Um, that's what I, that was my upbringing on cattle breeding. And so I am 100%, it's all about a EPDs and performance information. and so my first response to people that wanted to say, know, EPDs are bad, cattle. I can see them and breed them myself, was to say, you're wrong. Today I've probably fallen somewhere in between the two extremes between those that say, I don't need anything, but a computer to select the cattle I'm going to mate and breed and propagate to the people that, you know, on the other end of the spectrum. And maybe I'll just ask you, is cattle breeding, as far as breeders who are trying to produce more of the good ones for our customers and ourselves, is cattle breeding an art or is it a science?
Kelli:I think it's gotta a little bit of both, right? Matt? I mean, I think at the end of the day, you know, we have all these objective tools that we have, but I also think there is something to trying to specifically for people in different environments, right. Trying to match some of those cattle to those different environments, match them to your resources, match, you know, match your cow size to the resources that you have available to you for instance. That's always a really easy example, um, to be able to think about, I mean, I think it's as I think, animal breeding, right. Makes sense. Super super humbled very quickly. because at the end of the day, if I could figure out how to allow all of our members to control which half, right. Of the genetics that get passed down from parent to offspring each and every time, um, that would be, that'd be awesome. Right. But, uh, genetics can and animal breeding humbles everyone pretty fast. So no matter how much science you have, or no matter how much, um, even art you have or, or practical, experience you have, man, it can humble you pretty quickly when you, when you basically breed two animals together and those random halfs align and, and it gives you something that was totally unexpected.
Matt:well, it takes so long. I mean, we just, earlier in the podcast talked about the, the biology of, of the cow is such that, uh, we've gotta be pretty patient. And, and we don't get a very quick response to the selection decisions that or mating decisions that we made. I heard, mark Zuckerberg on a podcast a week or so ago when he was talking about feedback loop is just so long takes us so long with Facebook to know if, uh, fact checking thing... somebody was chastising him for the way they were throwing people off or regulating what they could or couldn't say about COVID or something on Facebook. And he said, you know, the feedback loop is just so long that it takes us awhile to know if our fact checking rules that we built have worked or have been too restrictive or whatever else. And I'm like, buddy, you have idea about a long feedback loop. Why don't you breed cattle? But it does. I mean, and that's the challenge and I think I agree with you. I think, I think it's both, I think there's room for all of us. And that's the cool thing about the Angus breed, um, is we've got so much diversity and I think I heard Jerry Connealy mentioned it in the, in the Angus podcast last week, we've got so much diversity in our pool. But I don't know if maybe sometime we've got more diversity in our association membership we even do in our genetics. Um, and, and that's not a bad thing. Um, we don't all have to do things the same. I mean, to me, that's been some of the challenge as a society that we've gotten to is that we feel like whatever party we affiliate with whatever church we go to or don't go to on a Sunday, whatever, whatever, um, we have to be all in and exactly like everybody that's around us. And the fact of the matter is not only is that not necessary, it's probably not healthy. And so, um, I look forward to whether somebody believes that EPDs are the end all be all, or whether they think it's. Invention in the history of the cattle industry. I, I think there is a pretty bright future for all of us, um, to find ways that we can find and select and propagate cattle that work for us and our customers. um, and understand that because of where we run cows or because of the way we market cattle or calves or bulls or whatever the case may be. We may not need the same set of EPD profiles. We may not put the same emphasis or the same selection criteria on ours and, and that's okay.
Kelli:Yeah. exactly. I, the diversity of this membership is a beautiful thing. I think we have a, just as big a bell curve, in the diversity of your membership, as I do in most of your, uh, traits that you collect. And so, um, yeah, I mean, it is, it's phenomenal. It's phenomenal that we get to serve this membership. And I think it, you know, the Angus association has done a great job of diversifying itself. So that way, you know, you have a, a media company. You know, that's how they aid in the success of producers is through really good advertising and putting them in the right spot. You have cab, which is, just a beast in itself and has done a lot of great things for, for Angus producers. And you have AGI and the way that we're approaching, right, trying to create value for Angus families is just, just try to characterize the genetics. And if somebody finds value in that, that's, that's great. But if somebody finds value in the association from a different focal point, you know, we're, we're okay with that too.
Matt:Well, you do a great job at it, and I'll go back to your bell curve analogy. You won't have to walk very far down the halls at American Angus to find people that will agree that I am several standard deviations away from the mean or norm. and, uh, I don't know that that's necessarily a, uh, flattering comment, but that's okay. that's uh, that diversity's all right All right. All right Sounds good. Well, Kelli, thank you very much for being with us today. Great information, and a lot to think about, and I'll, uh, I'll make sure and include associations, contact information, both online and the number there. Now I'm sure if anyone does have further questions and wants to get in touch with you, they can find your information there on that website and, and further the conference.
Kelli:Yeah, absolutely. Then one of my favorite things is, is helping people and being able to have these types of discussions. So appreciate you having me on Matt.
Matt:you bet. You bet. Thanks for being here, Kelly.
Kelli:Thank you.
Matt:Thanks for joining us for practically ranching, brought to you by Dalebanks Angus. If you enjoyed the podcast, heck even if you didn't... help us improve by leaving a comment with your review wherever you heard us. And if you want to listen again, click subscribe and catch us next week. God bless, and we look forward to visiting again soon.