Effect of SNP Quality Control LEVEL on Genomic Estimation of Relationships

No information about the quality control parameters of SNPs effects on the genomic relationship estimation with low dense panels is available. The number of alleles and the allele frequencies could modify the estimates of genomic breeding values. The objective was to evaluate the effect of quality control parameters on the portrayal of the cattle genome using a 9K Bovine Illumina chip.

All the information was provided by Holstein de Mexico A.C. Genotype information came from 458 genotypes of Holstein Cows obtained by a 9K Bovine Illumina chip. Only markers from chromosomes 1 through 29 were used (8,455 SNP´s). The genotyping rate of the cows, loss of markers and allele frequencies of the markers were tested using PLINK 1.07. The minimum genotyping rate allowed was 10% and 30%. The frequency for each marker, Hardy-Weinberg test, mendelian error and the IBD matrix detection were calculated using PLINK 1.07. The IBD matrices were compared through off diagonal means. Three genomics relationships matrices were built (Van Raden, 2008, Yang et al. 2010 and Forni et al 2011).

No differences were found among the means of the arrays and standard deviations (0.498, 0.499, 0.5, 0.5) respectively for 10% and 30% for each matrix, which suggest that we can use any of these three matrices without affecting quality control estimation of breeding values.

We acknowledge Holstein de Mexico A.C. and CGG-SAGARPA-México for sharing genomic data. This work received funding from CONACYT-México under the project “IDENTIFICACIÓN DE VARIACIONES ESTRUCTURALES Y REGIONES GENÓMICAS ASOCIADAS A RESISTENCIA A MASTITIS EN BOVINOS PRODUCTORES DE LECHE”(176039) and INIFAP through the Project “Incorporación de información genómica a los procesos de evaluación genética del ganado productor de leche” (1056821832)