Genome-wide association analyses of multiple traits in Duroc pigs using low-coverage whole-genome sequencing strategy

Ruifei Yang,Xiaoli Guo,Di Zhu,Cheng Bian,Yiqiang Zhao,Cheng Tan,Zhenfang Wu,Yuzhe Wang,Xiaoxiang Hu,Ning Li

DOI: https://doi.org/10.1101/754671

2019-01-01

Abstract:High-density markers discovered in large size samples are essential for mapping complex traits at the gene-level resolution for agricultural livestock and crops. However, the unavailability of large reference panels and array designs for a target population of agricultural species limits the improvement of array-based genotype imputation. Recent studies showed very low coverage sequencing (LCS) of a large number of individuals is a cost-effective approach to discover variations in much greater detail in association studies. Here, we performed cohort-wide whole-genome sequencing at an average depth of 0.73× and identified more than 11.3 M SNPs. We also evaluated the data set and performed genome-wide association analysis (GWAS) in 2885 Duroc boars. We compared two different pipelines and selected a proper method (BaseVar/STITCH) for LCS analyses and determined that sequencing of 1000 individuals with 0.2× depth is enough for identifying SNPs with high accuracy in this population. Of the seven association signals derived from the genome-wide association analysis of the LCS variants, which were associated with four economic traits, we found two QTLs with narrow intervals were possibly responsible for the teat number and back fat thickness traits and identified 7 missense variants in a single sequencing step. This strategy (BaseVar/STITCH) is generally applicable to any populations and any species which have no suitable reference panels. These findings show that the LCS strategy is a proper approach for the construction of new genetic resources to facilitate genome-wide association studies, fine mapping of QTLs, and genomic selection, and implicate that it can be widely used for agricultural animal breeding in the future.