Empirical Statistical Power for Testing Multilocus Genotypic Effects under Unbalanced Designs Using a Gibbs Sampler.

Area of Science:

• Animal genetics
• Quantitative genetics
• Statistical genomics

Background:

• Epistasis, or gene-gene interaction, significantly contributes to phenotypic variation in complex economic traits of animals.
• Traditional genetic association studies often overlook epistasis, potentially limiting the explanation of observed trait variations.

Purpose of the Study:

• To introduce a Bayesian method for inferring multilocus genotypic effects, accounting for epistasis.
• To evaluate the statistical power of this method under various unbalanced experimental designs through simulation.

Main Methods:

• A Bayesian approach utilizing a Gibbs sampler to estimate marginal posterior distributions of multilocus genotypic effects.
• Simulation studies with varied numbers of loci, within-genotype variance, and sample sizes in unbalanced designs.
• Estimation of mean empirical statistical power for testing combined genotype effects.

Main Results:

• The study provides empirical statistical power estimates for detecting epistatic effects under different unbalanced designs.
• A practical framework is demonstrated for calculating these power estimates for specific sample sizes.
• The developed method effectively handles complex genotypic interactions.

Conclusions:

• The empirical statistical power estimates are valuable for designing future genetic studies examining gene-gene interactions.
• Optimal experimental designs can be determined to maximize the detection of epistasis in complex traits.
• This Bayesian method enhances the ability to study the genetic architecture of complex economic traits in animals.