PROJECT FACT SHEET
Swine Cluster 4 (2023-2028)
Activity 16 | Animal Health
Enhancing Robustness in Pigs Using Advanced Genomics and Machine Learning.
Project Lead: Younes Miar, Dalhousie University
Status: Ongoing
Why is this project important?
In pigs, robustness is understood as the ability to combine a high production potential with resilience to stressors. Closely related to robustness, resilience in the pork sector has been described as the ability of animals to minimize the impact of environmental disturbances or to quickly recover from them. The two qualities go hand in hand, as breeding for resilience traits is considered a great opportunity for selection of more robust pigs in the modern pig industry.
Given the low margins in the Canadian pork sector and the significant investment of producers’ time and money in their animals, ensuring pig robustness is essential. The development of methods to select for more robust pigs is a challenging task, but a vital one for researchers. With that in mind, a key focus in breeding programs today is ensuring that pigs are productive and able to thrive in a variety of conditions. Much research has been conducted, and is ongoing, around adapting resilience indicators for genomic selection programs.
More and more, science is conferring with industry partners and academic institutions on the most effective means of enhancing robustness in Canadian pigs, just as they are currently doing with aspects like longevity, behavior and pork quality. In large part, the answer lies in cutting edge technologies like genomics and machine learning. The latter is a branch of artificial intelligence and computer science that focuses on using data and algorithms to enable AI to imitate the way that humans learn, gradually improving its accuracy.
What will researchers do?
Call on a vast array of experts from across the country, including researchers from the University of Guelph (U of G) and the University of Alberta (U of A), scientists from the Canadian Centre for Swine Improvement (CCSI), and the swine division of PIC.
Collect data on robustness and indicator traits such as survival and feeding behavior. They will then estimate the genetic parameters of these traits to determine if they are heritable and, if so, to what degree they are passed on to the next generation. Do they exhibit low, medium or high heritability?
Examine the relationship between resilience or robustness and other economically important traits. Based on this information, they will develop a profile of how these traits are controlled genetically.
Implement a genomic selection program through PIC using current methods employed in the pork industry. Researchers can then compare the effectiveness of such a program with that of machine learning and gauge whether the latter offers greater accuracy of selection to be implemented through PIC.
What will be the benefit of this research?
The project’s ultimate objective is to generate long term economic benefits for pork producers by developing robust and resilient pigs that are in high demand around the world. Pigs that are healthy and robust become more efficient, which serves as another boost to Canadian producers’ revenue, with an expected bump of $54 million in the three years following the project’s completion. In supporting the breeding of such pigs, this study addresses the growing public pressure for humane, sustainable production that is also environmentally friendly.
By the project’s end, scientists will have matched the set of genetic markers with robustness traits to derive the genomic breeding values for the selection of robust pigs. The selection tools they craft will be placed in the hands of breeding companies to provide more robust and healthier pigs to the commercial pig farmers of Canada. The results of this study will serve as a sustainable genomic solution to increase the survival, as well as the health and welfare, of pigs from weaning to market weight.
What has been done so far?
The project has recruited and trained four highly qualified personnel (HQP), including two PhD students (each for four years), and two Postdoctoral Fellows (PDF – each for two years). The postdocs will be responsible for modeling disease resilience and performing machine learning methods for genetic/genomic evaluation of robustness.
To date, researchers have estimated genetic and phenotypic (observable characteristics) parameters such as heritability and genetic correlations for many of the traits, and have genotyped approximately 16,000 animals.
Project status:
Currently in progress. Results expected in 2028.
Collaborators:
Mohsen Jafarikia, Canadian Centre for Swine Improvement (CCSI)
Graham Plastow, University of Alberta
Dan Tulpan, University of Guelph
Deborah Adewole, University of Saskatchewan
Duy Ngoc Do, Dalhousie University