# Efficiency of genomic and phenomic selection using mid-infrared milk spectra for milk production, somatic cell count, and udder type traits in French Lacaune dairy sheep

## TL;DR

This study compares using milk mid-infrared spectra to genomic data for predicting traits in dairy sheep, finding it more effective for some traits like milk production.

## Contribution

The study introduces phenomic selection using milk mid-infrared spectra as an alternative to genomic selection in dairy sheep breeding.

## Key findings

- Phenomic predictions using MIRS outperformed genomic predictions for milk production and udder health traits.
- Averaging lactation MIRS improved prediction accuracy compared to using a single spectrum.
- Combining MIRS with SNPs did not significantly improve prediction accuracy.

## Abstract

Summary: Animal breeding programs traditionally use phenotype and pedigree information to estimate breeding values of candidates to select for the traits to be improved. In dairy sheep, phenotypes are collected from ewes on-farm, and milk quality (protein and fat content) is measured from milk samples analyzed by midinfrared spectroscopy. Since 2015, the French Lacaune breed has switched to genomic selection, additionally using molecular information from genotypes to increase the accuracy of breeding values. This study tested a new approach called phenomic selection, using milk mid-infrared spectra (MIRS) instead of the molecular information to predict phenotypes used in national genomic evaluations. The results showed that using MIRS enabled more accurate prediction of phenotypes for milk production and udder health traits compared with molecular markers. However, this was not the case for udder type traits. The simultaneous use of molecular markers and MIRS data did not improve the phenotype predictions.

Summary: Animal breeding programs traditionally use phenotype and pedigree information to estimate breeding values of candidates to select for the traits to be improved. In dairy sheep, phenotypes are collected from ewes on-farm, and milk quality (protein and fat content) is measured from milk samples analyzed by midinfrared spectroscopy. Since 2015, the French Lacaune breed has switched to genomic selection, additionally using molecular information from genotypes to increase the accuracy of breeding values. This study tested a new approach called phenomic selection, using milk mid-infrared spectra (MIRS) instead of the molecular information to predict phenotypes used in national genomic evaluations. The results showed that using MIRS enabled more accurate prediction of phenotypes for milk production and udder health traits compared with molecular markers. However, this was not the case for udder type traits. The simultaneous use of molecular markers and MIRS data did not improve the phenotype predictions.

•Phenomic predictions with milk MIRS were higher than genomic predictions for milk production traits.•Averaging lactation MIRS improved phenomic predictions over using a single spectrum.•Random regression-best linear unbiased prediction and Bayesian reproducing kernel Hilbert space methods provided similar phenomic prediction accuracies.•Milk MIRS data preprocessing methods have no impact on phenomic predictions.•Adding milk MIRS to single nucleotide polymorphisms (SNPs) in prediction models did not improve phenotype predictions.

Phenomic predictions with milk MIRS were higher than genomic predictions for milk production traits.

Averaging lactation MIRS improved phenomic predictions over using a single spectrum.

Random regression-best linear unbiased prediction and Bayesian reproducing kernel Hilbert space methods provided similar phenomic prediction accuracies.

Milk MIRS data preprocessing methods have no impact on phenomic predictions.

Adding milk MIRS to single nucleotide polymorphisms (SNPs) in prediction models did not improve phenotype predictions.

Genomic selection uses molecular and pedigree information to accurately estimate genomic breeding values of animals from birth for traits in selection. Recent research in phenomic selection in plant production is opening up new opportunities in animal breeding. The approach of phenomic selection has been little studied in animal production. Here, we evaluate the efficiency of phenomic selection to estimate the phenomic values of phenotype-free females using mid-infrared spectral (MIRS) data from their milk samples. The phenotypes of 1,531 first-lactation French Lacaune dairy ewes were considered for traits included classically in the breeding goals, such as milk production and functional traits (SCS and udder type traits). The inclusion of standardized raw MIRS data instead of SNPs led to very low phenomic predictive abilities for udder type traits (Pearson correlations between phenotype and phenomic values from −0.08 to 0.07). For milk production traits, the phenomic predictions were superior to the genomic ones, in particular for lactation SCS (LSCS), with a predictive ability at 0.49 instead of 0.04. Overall, random regression-BLUP and Bayesian reproducing kernel Hilbert space methods gave equivalent results on phenomic predictions across all traits, with no impact from spectral data preprocessing. Finally, the efficiency of the combination of SNPs and milk MIRS in prediction models was low (average +3.8% for milk production and LSCS traits). Phenomic predictions could open up new prospects especially for the selection of nongenotyped females.