Comparison of kinship estimates in Santa Inês sheep using microsatellite and genome-wide SNP markers

ABSTRACT The Santa Ines breed stands out in meat sheep production in Brazil. However, this breed is under risk to lose important traits, especially due to pedigree errors and uncontrolled mating. The use of suitable molecular markers can help in the efficient parentage assignment and kinship analysis. In the present study, we aimed to evaluate the power of different numbers of microsatellite markers in comparison to a genome-wide SNP array in kinship estimation in Santa Ines sheep flocks. Blood samples from 257 animals raised in six flocks located in the Mid-North sub-region of Brazil were used for DNA extraction. The COANCESTRY program was used to propose the best kinship estimator among seven that were tested. The R package RELATED was used to obtain kinship estimates. The microsatellite data were obtained using four panels with different numbers of loci (10, 12, 15, and 19). The panel OvineSNP50 BeadChip (Illumina Inc.) was used for genotyping the animals with SNP data, which were used to calculate the genomic relationship matrix (G). The kinships obtained using SNPs and microsatellites were compared using descriptive analysis and Pearson correlation. It was possible to estimate relatedness between full siblings, half siblings, parent-offspring, and unrelated animals using the four microsatellite panels. However, panels containing less than 15 markers had lower efficiency. Most of the pairs of individuals were unrelated or half sibs. In the six flocks, several pairs of individuals showed some degree of relatedness. This indicates that a great number of related individuals have mated. Our results demonstrated that kinships estimated using microsatellite panels were highly correlated to those obtained using the SNP panel, with correlations above 0.89. Thus, a low number of microsatellite markers estimates kinship in Santa Ines sheep with similar efficiency to a genome-wide SNP panel.