Genomic predictions of milk coagulation properties

Among the characteristics that make milk optimal for cheese production, together with fat and protein, milk coagulation properties (MCP) must be included: rennet coagulation time (RCT), curd-firming time (k20) and curd firmness (a30). Moreover, an estimator of cheese yield, such as the individual laboratory cheese yield (ILCY), should be added. All these milk features are becoming influential in the breeding programmes oriented to cheese production, which is the main purpose for several livestock species or breed. The dataset was composed by 965 ewes; 769 of these were genotyped with Illumina Infinium Ovine SNP50 v1 BeadChip. A pedigree with more than six hundred animals was also available. All ewes had records for RCT and ILCY, while only 903 and 857 records were available for a30 and k20, respectively. In order to have a normal distribution of records, k20 was converted to its logarithm (logk20). Aim of this study was to estimate breeding values accuracy using two methods for variance component estimation: restricted maximum likelihood (REML) and Gibbs sampling (GIBBS). Using blupf90 family software, genomic breeding values (GEBV) were computed using a single-step approach. Accuracies (ACC) of GEBV were computed as the root square of reliability. The overall accuracy values for phenotyped ewes were 0.38±0.16 and 0.35±0.14 using REML and GIBBS, respectively. The highest ACC value was recorded for ILCY, with average values of 0.54±0.04 (REML) and 0.49±0.04 (GIBBS), followed by RCT that showed values of 0.44±0.04 (REML) and 0.37±0.04 (GIBBS), respectively. Strongly lower accuracies were observed for females without phenotypes: 0.11±0.07 for a30 and 0.06±0.04 for logk20. The most represented ram in the pedigree, with eleven daughters, showed an ACC average value of 0.62±0.10. The quite promising results obtained, especially for ILCY and RCT, suggest the possibility to include these traits as breeding goal in dairy livestock.