Genome−wide association study and genomic prediction of sow resilience based on reproductive traits

Robust animals, who are generally less susceptible to disease and require fewer medications, have greater resilience. As big data collection technologies have progressed, discovering new indicators of resilience by examining longitudinal data has become feasible. Environmental factor-induced variability in reproductive traits may affect an animal’s ability to adjust to changing environmental circumstances. However, no published research has used variations in reproductive traits to assess pig resilience. We here gathered data on reproductive traits of 11 321 Yorkshire, 3 317 Landrace, and 1 745 Duroc sows. Regarding reproductive traits, we computed each environmental standard deviation ( S D e ) as their resiliencies. The heritability of resilience traits ranged from 0.02 to 0.61 across three breeds. In total, 9 171 sows were conducted genotyping with three chips, namely, KPS® Zhongxin-1 Porcine Breeding ChipV2 (1 133 Yorkshire, 176 Landrace, and 116 Duroc sows), Chip_plus (2 815 Yorkshire, 1 017 Landrace, and 663 Duroc sows), and CAU50K (2 287 Yorkshire, 614 Landrace, and 350 Duroc sows). After quality control, 46 358 single nucleotide polymorphisms ( SNPs ) and 9 171 individuals were used for the subsequent genome-wide association study ( GWAS ) and genomic selection prediction. 130 significant SNPs and 3 259 candidate genes were found through GWAS on SD e of reproductive traits across three breeds. Using best linear unbiased prediction ( BLUP ), we performed breeding value estimation reliability for SD e of reproductive traits, and revealed that the reliability varied from 0.05 to 0.49. When genomic BLUP ( GBLUP ) and single-step GBLUP ( ssGBLUP ) were used, the average genomic prediction reliability increased by 0.02 (range: −0.02 to 0.09) and 0.04 (range: −0.01 to 0.08) for Yorkshire and Duroc sows, respectively. For SD e of ten reproductive traits and their original traits, the phenotypic correlations (|0.01| to |0.90|) and genetic correlations (|0.00| to |0.99|) were observed. Considering results comprehensively, SD e of litter weight, and SD e of survival rate could be used as indicators of resilience for all three breeds. Furthermore, SD e of number of mummies, SD e of number born weak, and SD e of number of stillbirth, could be as resilience traits for Yorkshire and Duroc sows, and SD e of number born healthy could be used for resilience evolution in Landrace sows. High-resilience individuals outperformed their low-resilience counterparts in terms of reproductive across three breeds. The foundation for integrating resilience into breeding procedures was established by these findings, which also broadened the range of molecular markers used to evaluate sow reproductive potential and disease resistance.