03069: Optimizing GBLUP and Machine Learning Genomic Breeding Values for Health and Performance Traits in Working Dogs
Grant Status: Open
Abstract
Working dogs are used to perform a variety of tasks including ones they were bred for or tasks that their natural abilities are adapted to perform. Successful working dogs require substantial investments of time and money as they are strategically bred, raised, and trained. Developing genomic breeding values to help inform breeding and management of working dogs will allow for more accurate selection and refined management of working dogs as knowledge of their potential for performance and health traits can be identified at birth through their DNA. This project will assess the genomic prediction model of GBLUP and four machine learning models for their prediction performance of 54 health and performance traits characterized in Seeing Eye guide dogs and personally owned sled dogs. It will identify the ability to predict traits in two very different management systems. After identifying the best performing model(s), investigators will produce a comprehensive set of breeding values for use. The over-arching goal of this project is to provide new tools for dog breeders to select for elite working dogs.
Publication(s)
Help Future Generations of Dogs
Participate in canine health research by providing samples or by enrolling in a clinical trial. Samples are needed from healthy dogs and dogs affected by specific diseases.