DESIGN OF THE STUDY
GOALS OF THE STUDY
The primary purpose of the Canine Behavioral Genetics Project is to identify and understand the genetic background of panic, anxiety related disorders, compulsive disorders, and aggression in dogs. It is possible, and likely in many cases, that such behavioral problems have a hereditary component. These types of behavioral problems are the reasons most commonly cited by owners who relinquish their dogs.
Our hope is that by identifying the genetic components of these behaviors, we can aid owners, handlers, trainers, and veterinary professionals in designing effective training and treatment protocols for affected dogs. In addition, if genetic screens were made available for behavioral problems, it would be possible to identify affected dogs when they are young, and implement these treatments early, when they will be most effective.
The secondary purpose of this project is to understand the nature of canine genetic diversity across the entire canine population (including both purebred and mixed-breed dogs) as well as population structure and diversity within and between pure breeds. This knowledge is important in order to properly structure inquiries into the history and distribution of canine disease. It is also crucial for answering questions about the history of bloodlines, populations, the development and relationships of breeds, and about the evolution of dogs in general.
These two goals are the reasons why we are looking for dogs that are both behaviorally affected and unaffected, and dogs that are both purebred and mixed-breed.
Finally, it is our hope that our research will provide insight into the nature and causes of some human psychiatric disorders. Dogs suffer from behavioral pathologies (i.e., separation anxiety, obsessive-compulsive disorders) that are similar in expression and probably biological origin to human disorders. Dogs are behaviorally and genetically more similar to humans than traditional model organisms (such as rodents) and offer other advantages while being observable in a more humane and naturalistic environment (owners' homes) than laboratory animals.
We are exploring the genetics of behavioral problems in dogs using a two-pronged research design.
First, we seek to identify families of purebred dogs in which behavioral problems, such as anxiety and/or aggression, are known to occur and perhaps can be traced using pedigrees.
Purebred dogs are particularly useful for this approach because mapping of traits, whether physical (coat color), disease related (cancer), or behavioral (anxiety, aggression), is often facilitated by using "genetically homogeneous" populations. This means that the shared genetic background of purebred dogs is circumscribed, or simply put, was handed down from a limited number of founder dogs for the breed. The resulting limited genetic variation within a given breed (also known as "founder effect") ensures that all or most of the dogs with a trait within a breed are more likely to share the same genetic variation or mutation. This dramatically simplifies the process of finding the genes for these traits.
An extension of this principle is that when we are examining single familes of dogs, these dogs are even more genetically similar (because they are closely related) to each other. This means that if we identify affected vs. unaffected individuals, it is easier to pinpoint where they differ genetically (than it would be in unrelated dogs, which may differ simply because they are unrelated) and therefore to narrow down the location of a disease gene.
In this fictitious family tree, there are a number of dogs (DOG_01, etc) for which we know something about their disease condition: some are carriers of the disease (half-filled), some have the disease (filled), while the remainder are disease-free (unfilled). From our DNA marker panels, let us say that we find a gene with strong association to a particular behavior in DOG_01, DOG_06, DOG_09, and DOG_11. Because these dogs are related, suggesting that the connection between the gene and the behavior is much more likely to be genetic than if we had collected them randomly without knowledge of their pedigrees.
This approach allows us to follow the connection between a gene and a behavior as it runs through a family with some certainty, which is a powerful way to find a gene for a behavior. For this reason, it is important to collect samples within families of verifiable lineage and relationships (i.e., pedigrees) in which the dogs are accessbiel, and owners are amenable to participating in studies of behavioral traits.
Association studies across unrelated dogs
Our survey unrelated dogs either within a breed or across breeds to identify individuals that are affected (cases) and unaffected (controls). Individuals who are unrelated are less genetically similar overall than related dogs would be, which makes their points of similarity more interesting if corollations can be shown with disease states. For example, a group of unrelated dogs that all exhibit a particular behavioral problem may also all exhibit a certain genetic mutation that is not found in unaffected dogs. Since they are unrelated, it is unlikely that this similarity is simply due to common ancestry, and this mutation then becomes a good candidate for investigation as a possible cause of disease.
For association studies, it is important to have samples within and across breeds that are large and comprehensive, for two reasons. The first is that samples must be large enough that we have a good chance of including both cases and controls. The second is that we must then have enough individuals in each category so that our statistical tests have adequate power. Our goal is to sample at least 50 dogs per breed. For some breeds, we are fortunate to already have many more samples than this, but our collection of others is still growing.