Prediction of selection response for threshold dichotomous traits

Area of Science:

• Quantitative genetics
• Animal breeding
• Statistical genetics

Background:

• Predicting response to selection is crucial in quantitative genetics.
• Classical formulas assume symmetry, which may not hold for dichotomous traits.
• Polygenic additive inheritance underlies many important traits.

Purpose of the Study:

• To present a novel formula for predicting the expected response to one generation of truncation selection.
• To account for asymmetry in response due to trait prevalence and selection rate.
• To provide a more accurate prediction model for dichotomous traits.

Main Methods:

• Utilized the threshold liability concept and normality assumption.
• Incorporated a bivariate normal integral term to handle asymmetry.
• Explained the relationship with the classical formula (R = iota rho sigma G) via Taylor expansion.

Main Results:

• The derived formula accurately predicts response, especially when selection rates deviate from 0.5.
• Demonstrated up to 15-20% departure from usual predictions at extreme selection rates (0.05 and 0.95).
• Illustrated properties with a sire selection example using progeny test data.

Conclusions:

• The new formula offers improved prediction accuracy for dichotomous traits under truncation selection.
• The model effectively addresses asymmetries not captured by classical methods.
• Applicable to various scenarios in animal breeding and quantitative genetics research.