Genomic accuracy in different genetic architecture and genomic structure

Abstract

Genomic selection has been widely implemented in national and international genetic evaluation in the animal industry, because of its potential advantages over traditional selection methods and the availability of commercial high density single nucleotide polymorphism (SNP) panels.Considerable uncertainty currently exists in determining which genome-wide evaluation method is the most appropriate. We hypothesize that genome-wide methods deal differently with the genetic architecture of quantitative traits and genomes. A genomic linear unbiased prediction method (GBLUP) and a genomic nonlinear Bayesian variable selection methods (BayesA and BayesB) were compared using stochastic simulation across three effective population sizes (Ne). Thereby, a genome with three chromosomes, 100 cM each was simulated. For each animal, a trait was simulated with heritability of 0.50, three different marker densities (1000, 2000 and 3000 markers) and number of QTL was assumed to be either 100, 200 or 300. Data were simulated with two different distributions for the QTL effect which were uniform and gamma (a= 1.66, b=0.4). Marker density, number of QTL and QTL effect distributions significantly affected the genomic accuracy with different Ne. BayesB produced estimates with higher accuracies in traits influenced by a low number of QTL, high marker density, gamma QTL effect distribution and with high Ne.