13 Feb Dr. Cothran’s Address to the IRMHA Membership

February 13, 2001

I’ve talked to this group on several occasions, and I’ve always found the most productive thing was to just give a few introductory remarks, and then to open the floor for questions. The kind of things Iêm used to talking about are generally not the kind of things horse breeders are particularly interested in, although some of the things I do are important to the horse breeders, I think, so what I would really like to do today is touch on three different things that I think are important for this registry and for some all of the registries in many respect, and many other types of domestic animals as well.

As Steve indicated, we made the change in 2000 from blood-typing to DNA testing, and anyone who has sent in DNA samples for DNA testing knows that it has not gone real smoothly. I could spend the rest of the evening telling you about the kinds of things that have gone wrong. There have been all kinds of problems that we had never anticipated in terms of starting to do DNA. There have also been some problems that we did anticipate, although they would not be nearly as bad if we hadn’t have had the unanticipated problems. The main problems we have seem to be over.

Let’s talk about why we made the change to DNA.I think that most everybody understands that the DNA typing gives us a little bit more power. Itês really not a big difference in terms of the ability to detect an incorrect parentage. With Rocky Mountain Horses, or really with almost any other breed, the blood typing gives us an ability of about 96 % to 98% to detect an incorrect parentage.

DNA testing or blood typing, or anything like that is a form of genetic marker testing. In the past we’ve done the blood typing where we look at blood groups and by chemical genetic polymorphisms, and this has been a very effective and powerful tool for a number of different things, but particularly for parentage verification and identification of individuals.

Blood Typing generally has an efficacy of about 96 to 98 percent. This means that for every 100 incorrect parentages examined, 96 to 98 will be detected. This is based upon testing 15 ® 17 systems.

Blood typing is looking at proteins, and these are gene products, and we are looking at something that is one step removed from the gene. It’s also something that does require blood, and that’s why we call it blood-typing for more than any other reason, in that it does require blood to do the testing.

DNA microsatellite typing has an efficacy in excess of 99 percent based upon testing 9 to 12 systems. The more systems tested the higher the efficacy, however the cost increase exceeds the gain in power above 12 systems.

DNA is the hereditary molecule itself. It is the bit of chemical in each cell that holds the information that codes for an individual. There are a lot of different ways to test DNA. The way we are testing is a kind of marker that is called a MSAT, or MicroSatellite. I won’t go into the details too much, but I do want to show you what a MicroSatellite is.

Types Of Genetic Marker Systems

Blood Typing ® Looks at proteins which are gene products.

Blood Groups ® Red Blood Cell Surface Antigens

Biochemical SystemsãSoluble structural proteins or enzymes

DeoxiriboNucleic Acid NDAãThe hereditary molecule. Includes genes & non-coding sequence.

RFLPãRestriction Fragment Length Polymorphism

MSATãMicroSatellite

SNPãSingle Neucleotide Polymorphism
The plate in the DNA where these letters represent the bits of information that are the genetic code, there are four different building blocks that make up the molecule that is DNA and we just normally use their alphabetic abbreviation for the chemical name for these building blocks, which are A, G, C and T.

The MicroSatellite is a place where two or more of these building blocks are repeated in sequence, so here we have AT, AT repeated ten times. DNA is a double stranded molecule where one strand is the mirror image of the other where A is always across from T, and G is always across from C.

The variation in these MicroSatellites is in how many times these repeats occur. We might have a situation where this represents one chromosome that came from the father and one chromosome came from the mother and you have a repeat of ten times in the father and twelve times in the mother, and the repeat is longer here. (The dots are just to offset the repeats). (See Table Two)

(Table Two)

Structure of a DNA MicroSatellite

Strand 1 (father)

AGCTTGCAGCTAGGCCTAG.ATATATATATATATATATAT.CGAATCGACTG
TCGAACGTCGATCCGGATC.TATATATATATATATATATA.GCTTAGCTGAC

Repeat Unit

AGCTTGCAGCTAGGCCTAG.ATATATATATATATATATATATAT.CGAATCGACTG
TCGAACGTCGATCCGGATC.TATATATATATATATATATATATA.GCTTAGCTGAC

Strand 2 (mother)

Now, this type of structure in the DNA molecule is a part of a region normally that has no function, and is what you may have heard of as the €junk DNAê, and it is non-coding DNA, because this repeat sequence doesn’t make any sense to the biological mechanisms that translate DNA into protein.

DNA or Blood Typing has great power for parentage determination. It has far less power for breed identification, although it can be useful in some cases.

There are two things about it that make it a powerful tool for parentage testing. One is that this repeat is difficult for the mechanism, the enzymes that duplicate DNA, to copy. It gets in there and it starts going ¯ATATAT©÷ likes it’s stuttering, and so when it is duplicating DNA there is a higher chance that it will make a mistake, and it will either put an extra AT in there or it will leave an AT out. Now, this doesn’t happen every time, but it happens far more frequently than mutations in coding DNA, or even just normal types of mutations of junk DNA because itês just subject to error. So, that makes these MicroSatellites highly variable because they are subject to more mutation, and mutation is what creates variation.

The other part of it is that it is junk DNA, non coding DNA, so if you have a mutation it doesn’t have any effect on the cellular function, and so you can accumulate variations more rapidly in a piece of junk DNA than you can in coding DNA, particularly DNA that codes for functional proteins like we have in blood typing. So that is the basis for the DNA tests that we do.

Genetic testing is very useful for determination of relationships among breeds. It also has use as means of determining genetic variation within a breed.

When we talk about the power of genetic testing, what we are really talking about if efficacy, which is the ability to detect an incorrect parentage. As I said, blood typing has an efficacy of about 96% to 98% based on testing about 17 systems that we routinely test for in our laboratory. When we go to DNA testing we have a higher efficacy because there is more variability, but, we are really only going up to something just above 99%, so we are not gaining a lot in power. Going to 98% to 99.5% is not a big increase in the power of the efficacy. We have 12 systems routinely, and if we get 12 systems in almost any breed of horse, our efficacy is 99.95% or greater. So, the probably of us detecting an incorrect parentage is almost 100%.

The real difference in going to DNA testing is that you don’t have to have blood. And as I think most everybody knows, what we are using as a sample for our DNA testing right now are pulled hair, and I know we donêt have a large number of people here but I would like to make one comment from the lab, something that we’ve just recently observed, and that is if you are pulling hair on your foal the instructions say pull mane or tail hair. On the foals, however, please pull the tail hair because you get a better root bulb and it’s easier to work with. There is DNA in both the mane hair and tail hair bulbs, it’s just easier to see in the tail hairs, it’s a little bit bigger and a little easier to handle. This needs to be emphasized to people when sending out kits and such. It really doesn’t make a lot of difference on grown horse, but it does make a difference on the foals.

So, this is essentially what we are doing with the DNA right now, and why DNA is going to be helpful.

There are a lot of other things that are going on with DNA in terms of genetic forces. I really won’t go into those tonight but if anybody has a question on things related to horse gene mapping I’ll be happy to try to answer those questions.

There have been a couple of other issues floating around, some of them for quite a while in this registry, and some others maybe just more recently, that we can look at with the genetic testing. One is breed identification. Both blood typing or DNA are of limited use in identifying an individual horse as to breed. It can be done but there are limitations, and the primary limitation is that you can’t just get a genetic type on a horse and say this is a Rocky Mountain Horse, or this is a Quarter Horse, or this is an Arabian Horse. The way you can use it is if you have the background information on your breed and you have a question about an individual horse as to whether it is Breed A or Breed B, then the genetic typing can give you with a very high percentage the ability to distinguish between the two breeds. There will be individuals that you can’t distinguish, and I like to call this the ¯Mr. Ed Syndrome, and that is to a large extent a horse is a horse, and when we look at genetic markers in horses, all breeds share a set of genetic markers. You can find them in any breed just about and I call these the generic horse markers. So, you can get an individual horse that essentially has a complete degeneric type and there is no way to tell what kind of horse that is from its genetic type. Of course, maybe you can look at it and tell, but in the laboratory there is no way to tell.

What the genetic testing is very good for is looking at genetic relationships among breeds, not individuals, but the whole breed compared to another breed, and it’s also useful for surveying genetic variability within the breed which has implications for small, rare breeds such as the Rocky Mountain Horse in basically trying to decide what the genetic health of the breed is.

So, what I want to do before I open up for questions is to just show a little bit of what we know about the Rocky Mountain Horse in terms of variability and relationship to other breeds.

Now this table has a lot of numbers in it, but lets just look at one column here.

This is the number of systems, and this is for blood typing, out of the seventeen, that are variable within an individual on average, and this is the proportion. So, for domestic horses the mean number of systems that are variable per individual within a breed is about 37% of them.

What I’ve done here is looked at just a few breeds kind of covering the range, with the Thoroughbred having among the lowest variability. There are some breeds with lower variability but they are breeds that are not so well known but just under 30% for the Thoroughbred.

Then we go down to something like the Peruvian Paso which is a breed with very high individual variability of about 45%.

The Rocky Mountain Horse is just a little bit below the average for domestic breeds at 35%. That’s pretty good, and that’s not anything to really worry about right now, but because it is below the average per 114 breeds that was tested in my laboratories, it is something that you want to keep an eye on a little bit. Itês the kind of thing that I know a lot of people are thinking about right now.

We’ve looked at variability in the breed in a number of different ways. One is looking at genetic variability through time. What this shows is the average genetic variations in the Rocky Mountain horse over roughly the last 25 years based upon age cohort. So this dot represents horses that were born in 1975, and this is 1990, and you see a lot of fluctuation. This is sampling error. This is not a lot of horses. This is the mean values at roughly 35%. What you really can see if that although we get year to year fluctuations, primarily due to sampling effect we are really not seeing a consistent trend for this to decrease very much. If you were to do statistics on this you might get a very slight downward trend but it’s not very high. We can look at some of these types of things, and at least up to this point the Rocky Mountain Horse is doing okay.

We’ve done this same sort of analysis in standardbreds, thoroughbreds and saddlebreds, and in thoroughbreds and saddlebreds you see exactly the same pattern except the fluctuations are much smaller because the sample sizes are much larger, but we basically see steady variation over that essentially same time period. In standardbreds we see a different pattern. We divide standardbreds into the trotty and pacing segments because of the way they breed them, that they are essentially different breeds even though there is a little bit of exchange between them.

In pacing standardbreds, over the period of about 1970 to 1990 there was a 5% decrease in genetic variability which is statistically significant. In trotting standardbreds over that exact same time period there was a 17% decrease in variability which is not only statistically significant but is quite likely biologically significant as well. We believe that is due to the breeding practices of using a very small number of stallions as the primary sires of each generation, and also then the sons of these stallions being the next generation of dominant breeders. Again, I don’t remember the exact number but something on the order of 6% of trotting stallions sired in excess of 60% of the foals in that breed, and that’s why you are seeing a big decrease of variability in that breed.

The last think I wanted to touch on before I open to questions was related to breed relationships, and the reason people have been interested in that relative to this registry lately is in terms of grade mare and grade horse registrations, the non-Rocky Mountain parent considerations on restricting which breeds might be allowed to be the parents of those, and I just wanted to show what we’ve got in that regard.

This is a tree showing degree of relationship among several horse breeds based upon blood typing markers again. I chose only those breeds that are what I call the North American gaited horse group that we have that on and that is basically this cluster. As you can see, the Rocky Mountain horse is in a group, and these are all very closely related in terms of genetic markers, that include the Kentucky Mountain, Mountain Pleasure and the American Saddlebred. So, these are the breeds, and we know this is nice conformation when things like this work. It is not a surprise but is in fact borne out by the data. People might be a little more surprised that the American Saddlebred is so close here, but if you think about the history of these breeds that is not quite as surprising. These are all horses that were developed from saddle stock that was in this area.

The rest of these are also very closely related, and if you put this into analysis where you have all 114 breeds that weêve tested, this grouping will essentially always hold up. This grouping is just some of the other breeds that are most closely related to this group which really are breeds relating to the Thoroughbred.

Now the issue of the Spanish horses, the Pasos and that sort of thing has been raised as well, so I took the Spanish horses and included them in the same tree, and just remember that the top one is the Rocky Mountain Horse, and then we get down here and have the Spanish breeds. The main point is that the Spanish breeds show a relationship that is less than the Thoroughbred related type breeds, and certainly less than the other gaited American horses. So, although there is probably contribution from these breeds, at least some of these types of horses to the Rocky Mountain horses and certainly to this group as a whole, currently they do not show quite as close a relationship.

I will be happy to entertain questions on any of these issues or of anything else that you may have a question about relating to genetics

Question #1: Is there any way you can pin down what markers might be more prominent in gaited horses than in your trotting horse.

Dr. Cothran: Well, one of the things that I briefly mentioned is the gene map, and the gene map of the horse is now at a point where it canêt be used functionally, and is being used to find markers for specific traits. ä As of right now, approximately 400 to 450 markers have been placed on the horse gene map of which at least 300 of those are variable enough to give us the opportunity that we could potentially find a marker associated with any particular trait, even multigenic traits. ä It would be multiple markers probably in that case, but even for quantitative type things we have the potential to find it.ä The way it would have to be done though is sometimes the tricky thing. ä The markers available, the way to test them is available, the cost is actually not too terribly high for that kind of thing, although it is not trivial. The way it would have to be done is that you have to have families where the traits you are interested in are segregating, basically that means that some of the offspring get the trait that you are interested in and some offspring don’t. When we talk about a ä family in this way we are talking about one or more half-sibling families. Basically what that means is that we’ve got one stallion that is passing on a trait to basically half of his offspring, and the other half of his offspring are getting the opposite of that trait, or are not getting that trait. We have to have something like that and that is difficult within a breed that every individual gaits. So what we would have to find would be an experimental population which would take some time to build up because we would probably have to cross some gaited horses to non-gaited horses to produce a stallion that could then be crossed to non-gaited mares to produce gaited and non-gaited offspring, assuming that it works that way. We don’t even know for sure that it works that way. A group that I talked to a couple of years ago and then I never heard anything again is the gaited Morgan. Iêm actually going out to the Morgan meeting and will be talking to them next Saturday in California, so I’m hoping that someone will bring that up again, but I don’t know who will be there related to that. ä But, it’s very possible to do that kind of study now.

Question #2: I have the blood results and DNA results that you did on a mare that I would like for you to look at. What we were trying to do is prove parentage on this horse and we just had the blood from the one parent of this mare, and I would just like for you to explain to me how that works and what percentage of that says that is her sire.ä Can you prove for that mare that one of her parents is a Rocky Mountain horse? ä Can you prove who her parent is?

Steve Autry: What we are going to talk about here is that in the past as an association we have used proof of parentage to the highest standard for determining whether parentage was correct, and we’ve used that as well in our grade mares in which we’ve required all three animals involved: sire, dam and offspring. We’ve always used blood typing in the past, and the question that this relates to is now that we have DNA, is there a difference between proof of parentage and proof of heritage, or whether they are from a particular breed of horse. Can you use either DNA to more selectively and effectively determine proof of parentage, or can you use a combination of blood and DNA to achieve essentially the same thing which is a high statistical likelihood that with only one parent you can say ¯that is a Rocky Mountain horse©÷.

Dr. Cothran: I talked about the efficacy, the ability to detect an incorrect parentage. What the converse of efficacy is, is that the probability that an individual who qualifies as the parent is the true parent is somewhere in that. So let’s say if we are blood typing for Rocky Mountain horses the efficacy is 97%. That means that there is a 3% chance that the parents that we looked at, one or the other or both are not actually the parent, but qualify by chance.ä Now, with DNA it is above 99%. When we only have one parent, that efficacy drops, and it drops at least 10%, maybe not quite so much with DNA but with blood typing it goes from about 97% to 87%, and these numbers are approximations. es”>ä With DNA, again I would say it goes from 99% to possibly a little higher. What that means is that the probability that the horse is not the parent is 3% or less if both parents are tested, and somewhere on the order of 10% if only the single parent is tested. That is a statistical number, and there are only a certain number of horses that could have bred to produce that foal, and in real life itês almost proof unless the other possible horses are close relatives like father, brother or son of the sire that you are looking at. In that case you are still almost certainly going to detect which one is correct if you have the stallion and the mare to compare. ä Now in your particular case we didn’t have the mare to look at. ä I can look at this type and say with a very high degree of confidence that this almost certainly is a Rocky Mountain horse. This horse has three of the markers that are very common in Rocky Mountain horses and are not common together in very many other breeds. They exist in other breeds but the probability of this not being at least half Rocky would be very, very small. But the thing is, with the ¯Mr. Ed©÷ think, you can breed a Rocky and a non-Rocky and get a horse that has Rocky markers, and the rest of the markers are generic. ä So, that is why you cannot prove what breed you have. All you can do is prove an individual is not the correct parent, but we only have a statistical probability that it is the correct parent. Again, the confidence level is very, very high, but it is not 100%.

Question (still relating to #2) In that particular case then, you said that one horse was her sire.

Dr. Cothran: Yes, and it fit with both blood typing and with DNA. Itês a very good fit, and the probability of that happening by chance with both sets of markers, which we’ve now looked at something well over 25 markers, that would be very unlikely that you would get that result by chance, even with just the single parent.

Question #3: If you would take 100 horse and go only with one parent, what would your chances of bringing into the breed horses that were totally non-Rocky mountain horses be if you use only one parent for parentage verification.

Dr. Cothran: You can introduce almost any breed of horse as the non-tested parent into the breed that way. We could pick up things that say this parent was not a Rocky by looking at variants that we would find in the foal that we have never seen in Rockies, or at least would have a very low probability of being found in Rockies. We could do that but that wouldn’t happen very often.

Question (still relating to #3) In looking at a population would you be able to tell if you were only testing for one parent, what percent of that population was really Rocky Mountain?

Dr. Cothran: No.

Steve Autry: Can you make a determination of proof of heritage to a higher degree of standard by the combination of DNA and blood together?

Dr. Cothran: Yes.

Steve Autry: What we are looking at is policy, issues, and our grade, and that it may change with the additional tool of DNA.

Dr. Cothran: Let me go back to this relationship issue which I guess is what you are talking about with the proof of heritage. It still needs a lot more work. The tree I showed you is based upon blood typing data and I have substantial blood typing data on over 100 different breeds of horses and over 50 different Mustang populations from North America and even some other places. We’ve only tested about 50 breeds with our DNA so far, and most of those are unusual, and we really don’t have much DNA data on Tennessee Walking Horses or a number of different breeds. In our lab we don’t have much data on Arabian horse from DNA at this point.ä We will, but we just simply haven’t gotten to it. We’ve been doing DNA testing for about five years now and it’s only been within the last year that we started doing large numbers. Most of the work that we had done in the past has been for a handful of smaller breed registries such as the Friesian and Lippizan registries and those horses arenêt too closely related to anything now. Now in trying to determine some of this relationship sort of thing, like you would generate that tree with, I have some questions about whether the DNA data is going to be more useful there or whether it may be less useful. My feeling at this point is that actually for the markers that are economically efficient for parentage testing is actually going to be less usefully for

looking at breed relationships. And the reason is what I explained earlier about all of this mutation range being higher. Although if you took the population back at some time in the past, you might be able to get some good ideas about relationships but over time these have been clouded a little bit by this higher mutation rate. ä And so, there are markers in blood typing that I have seen that are specific, maybe not to a single breed but at least to groups of a breed, so that we talk about markers that we see in the Rocky Mountain horse which indicate some Spanish ancestors. I havenêt seen any kind of markers like that in the DNA yet.ä You see markers that are found in a handful of breeds but not in another group of breeds but they don’t fit in a cluster. They could be a draft breed and a saddle

breed. Whereas with the blood typing we see better differentiations that forms these groups of breeds, and basically we see about six large groupings. ä One is the gaited North American horses, the second is the Thoroughbred related type horses and the third is the Spanish and the fourth would be the oriental Arabian type horses, and then we have the heavy draft horses and the true pony. ä Then there are a bunch of horse breeds like Friesians that donêt really fit well within any of those categories in terms of their genetic markers. Lipppizans donêt either but that is because they are a mixture of several different kinds of things. The question is still open in terms of that the DNA is going to test for.ä If we test for a large number of markers it will be better, but that is not something we do routinely because simply it is not economical to do that.

Question #4: With the use of both blood and DNA, if you were to take a known Rocky Mountain stallion or mare that is on file, and test it and the foal, without the other parent in laymanês terms what is the accuracy of saying that horse is out of that horse and has one Rocky Mountain parent, as opposed to using the highest degree.

Dr. Cothran: It would probably be around 95%, maybe 96% with both, if that is the correct parent, and the determination of the breed of that parent in terms of the markers is kind of hit or miss. But, if you know who that horse is, you know what its breeding is.

Steve Autry: So you can achieve higher accuracy with two techniques with only one parent, than you can with either or DNA or blood.

Dr. Cothran: Exactly. The power of these tests is based upon the variability of each system, but if you think about it like this, if you have one system that has a 50% chance of detecting an incorrect parentage and you test it, there you have a 50/50 chance. If you add another system that has a 50% chance of detecting incorrect parentage, then youêve got a 50% chance for the 50% you didn’t test, so basically you have a 75% chance. So you cut the probability of every system. With blood typing we have systems ranging from PI on the blood type where the probability for that system alone is about 55%, to some of the systems like the Keg?? Blood group where the probability may add 2%. With the DNA all of the systems are probably at least 30% individually, and some of them are probably 60% or higher. Every system you add cuts into that chance that you missed something.

Jerry LaFayette: So it’s safer to say that it indicates that one parent is a Rocky Mountain horse, that it more accurately states it if you do it with the two horses that horse is out of.

Dr. Cothran: What we test for is only to see if the parentage is correct, that is all we are looking at. ä We assume that if it is, then it is a Rocky Mountain horse because that is what is being sent to us.

Question #3: How much longer will blood typing be available.

Dr. Cothran: Actually as of this summer our laboratory and the Shelterwood Laboratory in Texas may be the only labs in North America still offering blood typing. I know that the University of California said as of July 1 they would not be doing blood typing. Steve Autry: If we wanted to continue blood typing in those situations where we were trying to prove identity for a heritage standpoint would blood typing still be available for a period of a reasonable length of time?

Dr. Cothran: I intend to continue to offer blood typing as long as we can.ä I use it as a research tool.ä We also have at least two or three registries that are not prepared to change to DNA yet and we are not taking the position that you either go with DNA or go with another laboratory, that is not the way we operate. ä We just signed a new 3 year contract with the Tennessee Walking Horses so we’ve got at least 3 years that we will continue blood typing, and we will probably offer it longer. There are a couple of reasons that we tend to. ä One is that I use it as a research tool, and I’ve got this tremendous database of 200 populations that if we quit doing it, I couldn’t use that information. Secondly the blood group testing has applications in jaundiced foal testing and NI testing. We think it is worth trying to maintain that capability for that type of testing. ä The reason that some of the laboratories are dropping blood typing is that you have to maintain horses to do blood typing and we currently are maintaining about sixty horses which I know you people know is not an inexpensive thing to do, and we’ve had as many as 90 that we’ve kept but we are cutting down on some of the production so we are cutting the numbers. But even just for NI testing we need to maintain about 25 horses just to have to compare the blood variations in a group that you can test to see if a mare is circulating antibodies.

Steve Autry: When you look at problems you have experienced from the field coming in to you, what advice other than to pluck from the tail do you have for our membership to expedite things so that we get a more rapid turn around of our material.

Dr. Cothran: The key think is just to make sure you get a good hair sample. You have to see that root bulb.ä It doesn’t look like a human follicle with a football shape, on a horse it looks slightly curled on the end, more like a comma on the end of the hair shaft. ä We need to have that bulb, if it is just hair it is not of any use to us. The main thing to remember is that the hair shaft itself is just protein, and is something that is produced by the cells but doesn’t have a nucleus and does not have DNA in it.

Question #6: How many hairs do you need?

Dr Cothran: We like to get at least 25 with bulbs. We actually use 3 but sometimes it doesn’t work so we have to get 3 more. Sometimes we get them and the follicles we get just aren’t good ones and that doesn’t work. We can test a single root bulb but our procedure is designed to test 3, and we want to have more to retest in case something doesn’t work. This DNA technology is a little more sensitive to error or laboratory failure, and that has been part of the problem with these frozen serum samples, as they don’t work nearly as well as we’d hoped they would and they donêt work as well as our initial work had indicated they would, but I think that is because when we initially started working with these frozen samples we were working with small numbers at a time and did repeated things to get the data into the laboratory. With these frozen samples we try to test all 12 systems at one time, for economic reasons, but what frequently happens is that about half of those don’t work the first time. Another thing that is tricky about DNA, and we knew this already but it’s showing up a little bit more than we’d thought, is a problem that when you do the test with blood typing you know immediately whether it worked properly. With the DNA you can get a product that is not correct but it isn’t obvious until you actually get to the parentage examination and then you have to go back and redo it, and actually you have to do both the foal and the parents because you don’t know which one is not correct. Sometimes you can tell immediately but not always.

Steve Autry: How long do you anticipate that those hairs will be good as far as future testing or retesting?

Dr. Cothran: A long time. We’ve tested some hair samples from a museum specimen about 80 years old that worked just fine.

Question #7: Since we are still a small breed, what should breeders be considering to prevent other medical problems from cropping up like ASD.

Dr. Cothran: I think the key thing is to continue to use a diversity of stallions. I donêt see that is a real big problem for a breed like this but the number one thing you have to worry about is having a small number of stallions dominating the breed.

Steve Autry: I think we have about 140 stallions that are doing most of the breeding for about 6500 horses. ä The vast numbers are coming from a small segment.

Question #8: When you talk about a form of line breeding, what percentage of blood can be on the stallion’s side and the mare’s side without having problems?

Dr. Cothran: Basically the way to look at it is that every generation you go back from an individual to a common ancestor you cut the contribution that that ancestor gave to that individual by half. If you have two individuals who share a parent they are related by 50%. If they share one grandparent they are related by 25%, which means that 25% of their genome is common, or shared, on average for each generation you go back. ä If you add two grandparents in common, or more than one shared ancestor you have to add that in. ää These genetic defects are generally single genes and every individual carries deleterious genes. The probability of them getting paired up in two individuals is very remote if the individuals are not related, but every degree of relationship you give increases the likelihood. I canêt give you exact percentages but essentially it is the inbreeding coefficient or what would be the probability of having two genes being identical by descent.ä However the probability that those are deleterious genes is not that and we don’t know what the proportion of deleterious genes in your population is. Now the way it kind of works is that in closed populations like a breed where you donêt have outside individuals coming in. Line breeding is a form of inbreeding, but it’s just where the common ancestor is more remote than what people think of as an incestuous type inbreeding, If you keep those common ancestors back around five generations or more, natural selection is going to get rid of a lot of those deleterious genes because natural selection and the breeder’s selection in the case of domestic animals, because if you see an animal that has a defect you are not going to breed it, or it will not be healthy enough to breed, or it will just die. If you have more rapid inbreeding where you move the common ancestor up in the pedigree, you increase the likelihood of pairing up something and spreading it around. ä The other thing is if you add new genes in you are continuing to add new deleterious genes in. ä They don’t become a problem until that individual’s genes are spread around widely, for example if it is a stallion. If that ever happens then you are introducing a new source, so you kind of have two competing things there. Outcrossing is good, in that you are continually adding new variations, but if you outcross a little but keep things closed you are actually adding some other potentially deleterious genes. It is a very complicated thing and it is hard to give you an exact answer.

Question #9: What is our inbreeding coefficiency right now:

Dr. Cothran: I can’t tell you for certain because we are waiting to get the entire studbook on computer. I did some preliminary estimates some years ago and I think we were looking at something on the order of 8%, but that was possibly a little bit higher that it is for the whole population because that was looking at the foundation animals and direct descendants of that group.

Question #10: How many Rocky Mountain horses have DNA identified and are on file at this time:

Dr. Cothran: I would guess that we probably have completed DNA testing on somewhere around 600 horses, it might be more. The big problem is right now is that there are a very large number that we have tested that we’ve had to test again and we are still testing again, and these are the frozen samples. The hair samples worked great. The new blood samples work great and donêt give us many problems. The old samples are being hard to try to get types on and we are having the same problems with everyone that we are doing it with.ä We are testing the Rockies and the Jockey Club and Thoroughbreds right now, but we are about to start testing the Saddlebreds and are also doing some other smaller registries. We know we are going to have similar problems. At least we know how to deal with those now to a better degree.

Steve Autry: A few years ago you were doing some research and trying to determine the position of the silver dapple gene. ä Are you doing any more work on that, or what is the status on that?

Dr. Cothran: I’m hoping that we are going to get back to that very soon. Shawn Phillips was doing that work had to stop that work but is now back. We have talked and think we have a very good chance. In some preliminary work she identified a candidate gene, which is a gene that we think is the gene that is responsible for the silver dapple color. She had some difficulty getting the sequence of the area that she needed. With DNA parts of that can be very easy to get the sequence on, and we have the training and equipment to do that. In fact, the instruments we use to do the DNA testing for parentage are actually designed for doing DNA sequencing. ä But, if the part of the gene that you are looking for happens to have more than about 50% made up of the G’s and C’s, that is technically more difficult to sequence, and it appears that the silver dapple gene has a high percentage of GCês which means that it is going to be harder to sequence than if it were just the normal stretch of DNA.

Steve Autry: Is that also the case with the Rockies as far as your ability to get them through the lab as compared to Thoroughbreds?

Dr. Cothran: Yes, we’ve had more problems getting DNA from frozen Rocky Mountain samples than we have from other breeds, and there is no explanation for that.

There were no more questions. We thank Dr Cothran for speaking to us and sharing his knowledge, and giving us this excellent information and explanation on the process of DNA analysis. We now have some very interesting things to think about. – Irene