Diallel analysis for sex-linked and maternal effects

Area of Science:

• Quantitative genetics
• Statistical genetics
• Animal breeding

Background:

• Genetic models incorporate sex-linked, maternal, and autosomal effects.
• Accurate estimation of genetic variance components and prediction of breeding values are crucial in quantitative genetics.

Purpose of the Study:

• Compare the efficiency of minimum norm quadratic unbiased estimation (MINQUE) and restricted maximum likelihood (REML) methods.
• Develop and evaluate an adjusted unbiased prediction (AUP) method for genetic effects.
• Assess the utility of the jackknife procedure for variance estimation and significance testing.

Main Methods:

• Monte Carlo simulations to compare MINQUE and REML estimation efficiencies.
• Development of the adjusted unbiased prediction (AUP) method.
• Application of the jackknife procedure for variance estimation and t-tests for significance.

Main Results:

• MINQUE(1) demonstrates comparable efficiency to MINQUE(θ) and offers advantages over REML in unbiasedness and computation.
• The AUP method provides easy computation and unbiased prediction of random genetic effects.
• Jackknife-based t-tests are effective for detecting significant variation.

Conclusions:

• MINQUE(1) and AUP are computationally efficient and statistically sound methods for genetic parameter estimation and prediction.
• The jackknife procedure is a reliable tool for assessing the precision of estimates and significance of variation.
• These methods are demonstrated effectively using mouse and silkworm data for variance estimation and genetic effect prediction.