Accuracy of genomic prediction for milk production traits in Mehsana buffalo.

Introduction: Genomic information can contribute significantly to the increase in the accuracy of genetic evaluation compared to relying solely on pedigree relationships. Hence, the objective of this study was to compare the accuracy of genomic prediction for 305 days milk yield (305DMY), 305 days fat yield (305DFY), 305 days solid-not-fat yield (305DSNFY) and 305 days protein yield (305DPY) traits in Mehsana buffalo using pedigree best linear unbiased prediction (PBLUP) and single-step genomic best linear unbiased prediction (ssGBLUP) models. Prediction bias was assessed by estimating the regression coefficient of corrected yield (Yc) on predicted breeding values (BVs). Methods: The phenotype dataset comprised of test day records of 10,897 Mehsana buffalo for milk yield, 10,896 for fat yield, 10,581 for SNF yield and 10,578 for protein yield were used for present study. A total of 4,107 blood samples of Mehsana buffaloes were collected and genotyped using BUFFCHIP 54K SNP array. After the quality control, final dataset comprised of 3,887 Mehsana buffalo with 53,292 SNPs. Results: The prediction accuracies obtained using PBLUP ranged from 0.078 to 0.088, whereas those from ssGBLUP ranged from 0.088 to 0.100. The average predictive accuracies across traits were 0.083 for PBLUP and 0.093 for ssGBLUP, representing an overall improvement of 11.37% with ssGBLUP. The regression coefficients of predictions ranged from 0.44 to 0.52 for PBLUP and from 0.45 to 0.61 for ssGBLUP across production traits. The average regression coefficients were 0.48 for PBLUP and 0.53 for ssGBLUP, indicating reduced prediction bias under the ssGBLUP model. Conclusion: ssGBLUP provides more accurate and less biased breeding value predictions than PBLUP for production traits in Mehsana buffaloes.