Best Prediction of the Additive Genomic Variance in Random-Effects Models

Area of Science:

• Quantitative Genetics
• Genomic Prediction
• Statistical Genetics

Background:

• Additive genomic variance is key in quantitative genetics but current models often ignore linkage disequilibrium (LD).
• Estimating genomic variance typically uses unconditional expectations, neglecting LD's contribution.
• This oversight can lead to inaccuracies in genomic prediction models.

Purpose of the Study:

• To introduce a novel best prediction (BP) approach for additive genomic variance.
• To incorporate the contribution of linkage disequilibrium (LD) into variance estimation.
• To provide a more accurate method for genomic prediction.

Main Methods:

• Developed a best prediction (BP) approach within the genomic best linear unbiased prediction (gBLUP) framework.
• Derived an empirical best predictor (eBP) for additive genomic variance.
• Compared the eBP performance against common estimation methods using genomic datasets.

Main Results:

• The novel BP approach provides the conditional expectation of additive genomic variance, integrating phenotypic data.
• This method explicitly accounts for the contribution of marker linkage disequilibrium (LD).
• The eBP demonstrated competitive or superior performance compared to existing methods.

Conclusions:

• The novel best prediction approach accurately estimates additive genomic variance by including LD.
• This method offers a more robust framework for genomic prediction, especially in the presence of LD.
• The empirical best predictor (eBP) is a promising tool for genetic variance estimation.