Aligning phenotypic and genetic trends: comparing trends from threshold and linear models in pigs

Area of Science:

• Animal Breeding and Genetics
• Quantitative Genetics
• Livestock Production

Background:

• Accurate estimation of genetic trends is crucial for effective animal selection.
• Genomic breeding values (GEBV) can be computed using linear or threshold models, depending on the trait's distribution.
• Transforming GEBV to a probability scale can aid in comparing across different trait types.

Purpose of the Study:

• To compare phenotypic and genetic trends using GEBV from linear and threshold models.
• To evaluate the effect of transforming GEBV to a probability scale on trend estimation.
• To assess the alignment of genetic and phenotypic trends for wean-to-finish mortality (WFM) and front leg structure (FLS).

Main Methods:

• GEBV were computed using linear and threshold models for WFM and FLS.
• Two methods were used to transform GEBV to probabilities from linear models.
• Genetic and phenotypic trends were calculated by averaging GEBV and phenotypes by birth year.

Main Results:

• Spearman correlations between raw and transformed GEBV were high (≥0.98), indicating minimal impact of transformation on selection candidates.
• Genetic trends derived from GEBV on the probability scale aligned well with phenotypic trends for FLS.
• For WFM, genetic and phenotypic trends showed less alignment, attributed to lower heritability, incidence, and environmental influences.

Conclusions:

• GEBV from linear models can be effectively transformed to the probability scale, showing strong alignment with GEBV from threshold models.
• The alignment of genetic and phenotypic trends is influenced by the chosen model, heritability, and trait incidence.
• Transformation to a probability scale aids in comparing genetic trends across traits with different distributions.