Linkage disequilibrium in Brazilian Santa In\xeas breed, Ovis aries

Amanda Botelho Alvarenga,Luís Fernando Batista Pinto,Luiz Lehmann Coutinho,Gota Morota,Gregori Alberto Rovadoscki,Gerson Barreto Mourão,Juliana Petrini,Gleidson Giordano Pinto Carvalho,Matthew L Spangler

doi:10.1038/s41598-018-27259-7

Amanda Botelho Alvarenga, Luís Fernando Batista Pinto + Show 7 more

Open Access

https://doi.org/10.1038/s41598-018-27259-7

Copy DOI

Abstract

For genomic selection to be successful, there must be sufficient linkage disequilibrium between the markers and the causal mutations. The objectives of this study were to evaluate the extent of LD in ovine using the Santa Inês breed and to infer the minimum number of markers required to reach reasonable prediction accuracy. In total, 38,168 SNPs and 395 samples were used. The mean LD between adjacent marker pairs measured by r2 and |D′| were 0.166 and 0.617, respectively. LD values between adjacent marker pairs ranged from 0.135 to 0.194 and from 0.568 to 0.650 for r2 for |D′| across all chromosomes. The average r2 between all pairwise SNPs on each chromosome was 0.018. SNPs separated by between 0.10 to 0.20 Mb had an estimated average r2 equal to 0.1033. The identified haplotype blocks consisted of 2 to 21 markers. Moreover, estimates of average coefficients of inbreeding and effective population size were 0.04 and 96, respectively. LD estimated in this study was lower than that reported in other species and was characterized by short haplotype blocks. Our results suggest that the use of a higher density SNP panel is recommended for the implementation of genomic selection in the Santa Inês breed.

Highlights

Genomic information is currently used in animal breeding programs to enable selection for difficult to measure traits, increase the overall rate of genetic gain, and to improve the understanding of genetic and biological causes underlying phenotypic variation
Through the knowledge of the degree of linkage disequilibrium (LD), it is possible to define the density of genetic markers necessary to achieve a certain accuracy of prediction and to determine when the estimates of genetic marker effects should be updated
Increased marker density can dilute individual marker effects if, for example, two markers are associated with the same QTL and the two markers are in high LD with each other

Summary

Introduction

Genomic information is currently used in animal breeding programs to enable selection for difficult to measure traits, increase the overall rate of genetic gain, and to improve the understanding of genetic and biological causes underlying phenotypic variation. Genomic selection (GS) is an approach which uses genome-wide markers simultaneously to predict breeding values[1] This approach has been shown to increase the rate of genetic gain when pedigree-based selection is suboptimal[1], which is the case for lowly heritable traits. The extent of LD can vary between and within species due to evolutionary history and population structure mainly characterized by insertions, deletions, chromosomal rearrangements, or inversions[4] This association between markers and causal mutations may change overtime due to recombination and selection[4] necessitating the re-estimation of marker effects. Beynon et al.[16] mentioned the importance of studies focused on breeds as a chance to identify variation and understand the biological mechanisms that enable these breeds to survive in different local environments

Objectives

Methods

Results

Conclusion