Optimising within and between population genomic prediction in aquaculture breeding

Lead Research Organisation: University of Edinburgh
Department Name: Royal (Dick) School of Veterinary Scienc

Abstract

Aquaculture is the fastest growing agricultural sector world-wide and in the UK and Scotland in particular salmon is the most important farmed species. The relatively recent domestication of aquacultural species including salmon produces both significant challenges and substantial opportunities for the use of breeding to improve both production and welfare. The recent development of single-nucleotide polymorphism (SNP) genotyping arrays raises the possibility of substantially increasing rates of genetic improvement using genomic selection approaches. Genomic selection utilises computational analysis to combine genome-wide SNP data with trait information to identify fish carrying the best of the naturally occurring genetic variation for desirable characteristics and hence select the best fish to breed future generations.

The project is a collaboration with the Salmon breeding company Landcatch Natural Selection (LNS), part of the Hendrix Genetics breeding company. LNS has been using genomic breeding approaches for several years and was the first to identify (in collaboration with Roslin Institute and others) and utilise a gene associated with increased resistance to disease in their breeding programme. The company will provide data from several different populations of salmon that have been recorded for traits including parasite resistance and have been typed using a genome-wide SNP array. The objective of the project is to explore the efficiency of genomic prediction within and between populations using the real data and data simulated according to alternative genetic models. The main task will be to optimise the accuracy of selection by adjusting the emphasis put on within and between population genomic information and to suggest additional information that might be collected to further enhance accuracy.

Studentship Projects

Project Reference Relationship Related To Start End Student Name
BB/P504993/1 01/10/2016 30/03/2021
1804817 Studentship BB/P504993/1 01/09/2016 31/08/2020
 
Description Taking under consideration that I am in the middle of my research project, I have explored the impact of population structure in a breeding program design and I have simulate a breeding program with the aim to optimize genomic prediction in aquaculture. In order to achieve that, I had to explore the structure of commercial salmon populations from Landcatch (Industrial partner - Hendrix Genetics group). The representation of Landcatch population structure was used to simulate a breeding program that reflects the existing salmon breeding populations. The aim is to investigate the impact in genomic prediction of mixing individuals between parallel year groups in Atlantic salmon breeding programs.
The objective of Landcatch dataset characterization is to estimate the genetic diversity levels within and between the breeding populations. A principal components analysis (PCA), based on genotype data, shows a distinct differentiation between the four breeding lines. The levels of genetic diversity within each population was similar among all populations and there were no significant levels of inbreeding within each population. The average genetic distance between populations was calculate and compared with the simulated populations. Further analysis in 6 different breeding populations (Scotland and Chile) reveal many similarities between the Landcatch and simulated populations. Observed differences between the Scottish and Chilean populations maybe they are due to the population structure of each breeding program, the selection pressure or differences on the sample size. Moreover, similarities (high correlations) suggest a higher relatedness and also a possible higher accuracy of prediction.
An aquaculture breeding program was simulated with aim to reflect the existing salmon breeding populations and explore the impact of gene flow in the accuracy of genomic prediction within and between the breeding lines. Once the base population was created, we simulate the four parallel lines for 10 discrete generations under random selection. The process was replicated 20 times. Genetic distance and accuracy of genomic breeding values were used as criteria of comparison. The results show that with no mixing the genetic distance increased between populations and the genetic variation within population decreased. There was no increase of accuracy by combining data across populations. Even a small percentage of mixing decreased the genetic distance between populations and increased the genetic variation within population. The higher the percentage of mixing the faster the lines became similar. The between accuracy of prediction climbed as the percentage of mixing increased. The increase in accuracy from the combined evaluation approach compared to the within evaluation approach was greater with an increased percentage of mixing.
Exploitation Route Although the award is still in progress there are many applications.
In academics the characterization of a salmon breeding program by estimating the genetic diversity levels within and between the populations will help scientists to understand the structure and evaluate possible genetic effects on desirable traits in future breeding and selection programs. A representation by simulation of a salmon breeding program will allow the investigation of different scenarios like the impact of gene flow on genomic selection when individuals mixed between the populations. Moreover, the estimation of genomic prediction reliability across populations using genotype data will be a new method to estimate the expected reliability on target population based only on genotype information.
In non-academics breeding companies will be able to reduce costs of annual tests by reducing the need for animals. An estimation of genomic prediction reliability will help them to estimate the accuracy of predicting breeding values based on the existing genotyped populations without the need to repeat annually tests or challenges. Additionally, the simulation method will give breeders a better image of the current and future population structure in order to adjust their strategy and improve the efficiency of genomic prediction.
Sectors Agriculture, Food and Drink
 
Description Because my award is still in progress there are not any use of my findings.
First Year Of Impact 2019
Sector Agriculture, Food and Drink
Impact Types Economic
 
Description BBSRC KTN CASE Studentship
Amount £80,000 (GBP)
Organisation Biotechnology and Biological Sciences Research Council (BBSRC) 
Sector Public
Country United Kingdom
Start 09/2016 
End 08/2021
 
Description Industrial Partner 
Organisation Landcatch Natural Selection Ltd
Country United Kingdom 
Sector Private 
PI Contribution Our contribution is the characterization of the population genetic structure of commercial salmon populations, in order to explore the efficiency of genomic prediction within and between populations using the real data and data simulated according to alternative genetic models.
Collaborator Contribution The animal breeding company (Landcatch) will provide data from several different populations of salmon that have been recorded for traits including parasite resistance and have been typed using a genome-wide SNP array.
Impact Taking under consideration that the award is still in progress, we have characterize the population genetic structure of commercial salmon populations. Moreover, we have create a simulation that reflect the exist real populations in order to explore the impact of gene flow in the accuracy of genomic prediction.
Start Year 2016