Application of single step genomic BLUP under different uncertain paternity scenarios using simulated data

Area of Science:

• Animal Genetics and Breeding
• Genomic Selection
• Quantitative Genetics

Background:

• Accurate genetic evaluations are crucial for beef cattle breeding programs.
• Paternity uncertainty in pedigree data can reduce the accuracy of genetic predictions.
• Genomic selection models aim to improve prediction accuracy by incorporating marker information.

Purpose of the Study:

• To evaluate the performance of BLUP and ssGBLUP models under varying degrees of paternity uncertainty.
• To assess different strategies for scaling the genomic relationship matrix (G) to the pedigree relationship matrix (A22).
• To investigate the impact of paternity uncertainty on the accuracy, bias, and inflation of estimated breeding values (EBVs).

Main Methods:

• Simulated beef cattle data including genotypes, pedigree, and phenotypes for age at first calving (AFC) and weight at 550 days (W550).
• Implementation of BLUP (phenotypic and pedigree data) and ssGBLUP (phenotypic, pedigree, and genomic data) models.
• Creation of paternity uncertainty scenarios with 0%, 25%, 50%, 75%, and 100% multiple sires (MS).
• Testing four G matrix scaling strategies to match the A22 matrix.
• Evaluation of accuracy, bias, and inflation across different animal groups (ALL, BULL, GEN, FEM, YOUNG).

Main Results:

• BLUP model accuracies decreased significantly with increasing paternity uncertainty, particularly for genotyped and young animals (up to 87.8% reduction).
• ssGBLUP model accuracies also decreased with paternity uncertainty, but the reduction was substantially less than with BLUP (38-44.6% reduction).
• G matrix scaling strategies did not significantly impact results for ALL, BULL, and FEM groups, indicating robustness to missing pedigree information in these groups.

Conclusions:

• The ssGBLUP model provides more reliable and less biased EBVs than BLUP, especially for young animals with incomplete pedigree information and paternity uncertainty.
• Ensuring compatibility between the G and A22 matrices is essential for accurate and unbiased genomic predictions using ssGBLUP, even with missing pedigree data.
• ssGBLUP is a valuable tool for improving genetic evaluations in beef cattle populations facing challenges with pedigree accuracy.