Predicting bull fertility using genomic data and biological information

Area of Science:

• Animal Genetics
• Dairy Science
• Genomic Prediction

Background:

• Fertility traits are crucial for the dairy industry, yet genomic prediction has largely focused on cows, neglecting bull fertility.
• Sire Conception Rate (SCR) is a key indicator of bull fertility, vital for genetic improvement programs.

Purpose of the Study:

• To assess the feasibility of genomic prediction for Sire Conception Rate (SCR) in US Holstein dairy bulls.
• To evaluate the impact of incorporating functional genomic information into predictive models for enhanced accuracy.

Main Methods:

• Kernel-based genomic prediction models were used, analyzing all single nucleotide polymorphisms (SNPs) or subsets of functionally relevant markers.
• Both single- and multi-kernel models with linear and Gaussian kernels were tested.
• Predictive ability was assessed using 5-fold cross-validation.

Main Results:

• The entire set of SNPs yielded predictive correlations around 0.35.
• Kernel models incorporating significant SNPs achieved the highest accuracy, improving predictions by up to 5% over standard whole-genome approaches.
• Gaussian kernels outperformed linear kernels, irrespective of the SNP set used.

Conclusions:

• Genomic prediction of bull fertility is achievable in dairy cattle, enabling informed genetic decisions.
• Utilizing Gaussian kernels and relevant genetic markers offers a promising alternative to standard genomic prediction methods.
• Further research into integrating gene set information into prediction models is warranted.