Multiple-trait genomic evaluation for milk yield and milk quality traits using genomic and phenotypic data in buffalo in Brazil.

The objective of this study was to compare the multi-trait model using pedigree information and a model using genomic information in addition to pedigree information. We used data from 5896 lactations of 2021 buffalo cows, of which 384 were genotyped using the Illumina Infinium(®) bovine HD BeadChip, considering seven traits related to milk yield (MY305), fat (FY305), protein (PY305), and lactose (LY305), percentages of fat (%F) and protein (%P), and somatic cell score (SCS). We carried out two analyses, one using phenotype and pedigree information (matrix A) and the other using the relationship matrix based on pedigree and genomics information (a single step, matrix H). The (co)variance components were estimated using multiple-trait analysis by the Bayesian inference method. The model included the fixed effects of contemporary groups (herd-year and calving season), and the age of cow at calving as (co)variables (quadratic and linear effect). The additive genetic, permanent environmental, and residual effects were included as random effects in the model. The estimates of heritability using matrix A were 0.25, 0.22, 0.26, 0.25, 0.37, 0.42, and 0.17, while using matrix H the heritability values were 0.25, 0.24, 0.26, 0.26, 0.38, 0.47, and 0.18 for MY305, FY305, PY305, LY305, %F, %P, and SCS, respectively. The estimates of breeding values in the two analyses were similar for the traits studied, but the accuracies were greater when using matrix H (higher than 8% in the traits studied). Therefore, the use of genomic information in the analyses improved the accuracy.