By Carol Beuchat PhD
The pedigree is the genetic history of a dog. The pedigree database of a breed is the genetic history of the breed. For breeds that began with a group of founder dogs and since have maintained a closed stud book, knowing the relationships among dogs in a breed can allow you to extract information about how the gene pool has changed over the generations, and it can also reveal the current "genetic structure" of the breed.
If you think of all of the animals in your breed as spread out across a landscape, but organized in clusters of genetically similar animals, you can envision of the genetic structure of your breed as the arrangement of these genetic clusters and their distances from each other. For instance, we might imagine all of the living Golden Retrievers on a landscape with two clusters, one that will be mostly bench-bred dogs, and another with dogs from field lines. We could zoom in on the bench line cluster and see that it is divided perhaps into groups of dogs in different countries, with the American dogs as a big cluster, and the UK dogs as a cluster that might also include a lot of dogs from other countries with strong British influence. Zooming in again, we might be able to distinguish clusters of genetically similar dogs that reflect particular kennels that have been especially influential and productive. As you zoom in and scan around, you can also see smaller and smaller clusters, and if you go far enough you could see the clusters of all offspring produced by a particular sire and dam.
These clusters of genetically similar dogs are the consequence of how we breed dogs, and in fact they are typical of all domestic animals. In purebred dogs, breeding is not random; if it was, existing clusters would become less and less distinct over time until eventually all of the dogs would be in one big clump. Instead, we tend to breed dogs that are related because they share particular traits that we want to produce in the offspring. Over generations of breeding within a group of related dogs and selecting for the desired traits, the dogs in the group become more and more similar genetically, and their similarity to dogs in other groups decreases. When breeders talk about "outcrossing", they are referring to breedings between dogs that are in different clusters. The more dissimilar the clusters are genetically, the greater the degree of outcrossing.
If we know there is genetic structure in a population of animals such as a dog breed, we can use that information in a very clever way. Let's say you have a colony of Labrador Retrievers that are being bred to produce service dogs. There has been an increasing problem of elbow dysplasia in the dogs, which of course is very undesirable in a dog headed for a career where its ability to perform particular duties is essential. Elbow dysplasia appears to be genetic, but despite decades of genetic research, the genes that cause it remain a mystery. So now you have a serious genetic issue in your breeding stock and no DNA test to screen for it. It would seem that you have little choice but to remove the affected dogs from the breeding program, and possibly also their parents, siblings, and offspring, aunts and uncles - and how far should you go? Clearly, this is going to adversely affect the size and diversity of your gene pool. Is there a better way?
If you have the pedigree records for your dogs, you can use that information to reconstruct the genetic landscape for your breeding colony. With genetically similar dogs grouped together, maybe you can identify clusters that have more than their fair share of dogs affected by elbow dysplasia.
In fact, this is exactly what happened to the Royal Dutch Guide Dog for the Blind Association (Ubbink et al 1998). They found that 17% of the 250+ dogs in their colony had elbow dysplasia. To develop a breeding plan that would reduce the incidence of elbow dysplasia in their dogs while minimizing damage to their gene pool, they used their pedigree information to reconstruct the genetic landscape of their breeding dogs, clustering together dogs based on genetic similarity. After they established the clusters, they superimposed on this landscape the locations of all of the dogs that had been diagnosed with elbow dysplasia.
This is a "dendrogram" of the genetic relationships of the dogs in the Dutch breeding colony. A dendrogram is a tree-like diagram that clusters dogs by degree of relatedness, then connects the clusters with branches that reflect the relationships among them. (If you're not familiar with dendrograms, you can learn how to read them HERE.)
The pedigree is the genetic history of a dog. The pedigree database of a breed is the genetic history of the breed. For breeds that began with a group of founder dogs and since have maintained a closed stud book, knowing the relationships among dogs in a breed can allow you to extract information about how the gene pool has changed over the generations, and it can also reveal the current "genetic structure" of the breed.
If you think of all of the animals in your breed as spread out across a landscape, but organized in clusters of genetically similar animals, you can envision of the genetic structure of your breed as the arrangement of these genetic clusters and their distances from each other. For instance, we might imagine all of the living Golden Retrievers on a landscape with two clusters, one that will be mostly bench-bred dogs, and another with dogs from field lines. We could zoom in on the bench line cluster and see that it is divided perhaps into groups of dogs in different countries, with the American dogs as a big cluster, and the UK dogs as a cluster that might also include a lot of dogs from other countries with strong British influence. Zooming in again, we might be able to distinguish clusters of genetically similar dogs that reflect particular kennels that have been especially influential and productive. As you zoom in and scan around, you can also see smaller and smaller clusters, and if you go far enough you could see the clusters of all offspring produced by a particular sire and dam.
These clusters of genetically similar dogs are the consequence of how we breed dogs, and in fact they are typical of all domestic animals. In purebred dogs, breeding is not random; if it was, existing clusters would become less and less distinct over time until eventually all of the dogs would be in one big clump. Instead, we tend to breed dogs that are related because they share particular traits that we want to produce in the offspring. Over generations of breeding within a group of related dogs and selecting for the desired traits, the dogs in the group become more and more similar genetically, and their similarity to dogs in other groups decreases. When breeders talk about "outcrossing", they are referring to breedings between dogs that are in different clusters. The more dissimilar the clusters are genetically, the greater the degree of outcrossing.
If we know there is genetic structure in a population of animals such as a dog breed, we can use that information in a very clever way. Let's say you have a colony of Labrador Retrievers that are being bred to produce service dogs. There has been an increasing problem of elbow dysplasia in the dogs, which of course is very undesirable in a dog headed for a career where its ability to perform particular duties is essential. Elbow dysplasia appears to be genetic, but despite decades of genetic research, the genes that cause it remain a mystery. So now you have a serious genetic issue in your breeding stock and no DNA test to screen for it. It would seem that you have little choice but to remove the affected dogs from the breeding program, and possibly also their parents, siblings, and offspring, aunts and uncles - and how far should you go? Clearly, this is going to adversely affect the size and diversity of your gene pool. Is there a better way?
If you have the pedigree records for your dogs, you can use that information to reconstruct the genetic landscape for your breeding colony. With genetically similar dogs grouped together, maybe you can identify clusters that have more than their fair share of dogs affected by elbow dysplasia.
In fact, this is exactly what happened to the Royal Dutch Guide Dog for the Blind Association (Ubbink et al 1998). They found that 17% of the 250+ dogs in their colony had elbow dysplasia. To develop a breeding plan that would reduce the incidence of elbow dysplasia in their dogs while minimizing damage to their gene pool, they used their pedigree information to reconstruct the genetic landscape of their breeding dogs, clustering together dogs based on genetic similarity. After they established the clusters, they superimposed on this landscape the locations of all of the dogs that had been diagnosed with elbow dysplasia.
This is a "dendrogram" of the genetic relationships of the dogs in the Dutch breeding colony. A dendrogram is a tree-like diagram that clusters dogs by degree of relatedness, then connects the clusters with branches that reflect the relationships among them. (If you're not familiar with dendrograms, you can learn how to read them HERE.)
This dendrogram instantly told them is that all of the cases of elbow dysplasia in their colony were occurring in a group of related dogs indicated by the partially shaded bars at the bottom of the tree. They could assume that all of the dogs in the groups with affected animals are a risk for elbow dysplasia because they have a high degree of genetic similarity. Removing just the dogs in those groups from the breeding program should be effective in reducing the incidence of elbow dysplasia in their colony.
Think about how useful this would be in the management of a genetic issue in your breed. Instead of shuffling through hundreds of pedigrees looking for patterns, instead of trying to figure out the mode of inheritance of a complicated trait, instead of spending a fortune on research to locate the genes causing the problems, you can home in on the likely genetic source using a diagram of the genetic structure of your breed. You don't need to know the genes involved. You don't need to know mode of inheritance. You don't need to spend thousands and wait years in the hope that geneticists will get lucky and find a single gene that you can test for.
Furthermore, you can do this with ANY trait that has a genetic basis. It can be a disease, a behavioral trait, a point of conformation, pretty much anything. And in fact, the more heritable the trait is, the more clearly it should be revealed in specific clusters on the genetic landscape.
And think about this. If you had a dendrogram of all of the currently reproductive dogs in your breed, and if people reported information about health issues in their dogs when they were diagnosed, the clusters containing affected dogs would become apparent as the reports came in. You would be able to identify emerging problems or developing patterns long before they became significant enough for breeders to realize there was an issue. Furthermore, reporting doesn't even need to be complete. More data would definitely be better, but missing data won't prevent you from figuring out where in the breed a problem is coming from. And if it should be necessary to do any DNA analyses, you would know where to find the dogs that are most and least likely to carry the genes of interest, which would make it much easier and less expensive to nail down the problem.
Think about how useful this would be. With an up-to-date pedigree database, you could produce a diagram of the genetic landscape of your breed that would help you nip genetic problems in the bud, and it would also be useful to identify where to find dogs to breed to that are unlikely to carry the genes you want to avoid. It could also help you identify groups of dogs that would be appropriate for outcrossing because the diagram will reveal the clusters that are less similar genetically to the one your dog is in. In fact, breeders of Icelandic Sheepdogs were able to identify the dogs in their small breed that were most valuable genetically because they carried genes that were rare in the breed, and from this they could develop breeding strategies that would minimize the loss of valuable genetic diversity and make the best use of the diversity remaining in the breed.
Trying to navigate through the gene pool of your breed without a picture of the genetic landscape, avoiding the genetic landlines you know about and keeping fingers crossed that you won't discover the ones you don't, is like trying to find an address in New York City without a map. The information you need to make your journey safer, easier, and less expensive is available to you now, in the pedigrees that record the genetic history of your breed.
- G. J. Ubbink, J. van de Broek, H. A. W. Hazewinkel, J. Rothuizen. 1998. Cluster analysis of the genetic heterogeneity and disease distributions in purebred dog populations. Vet Rec 142: 209-213.
Learn more about the genetics of dogs in ICB's Online Courses.
*** Population Genetics for Dog Breeders ***
Next class starts 30 March
Visit our Facebook Groups
ICB Institute of Canine Biology
...the latest canine news and research
ICB Breeding for the Future
...the science of animal breeding
*** Population Genetics for Dog Breeders ***
Next class starts 30 March
Visit our Facebook Groups
ICB Institute of Canine Biology
...the latest canine news and research
ICB Breeding for the Future
...the science of animal breeding
