Response from marker-assisted selection when various proportions of animals are marker typed: a multiple trait simulation study relevant to the sheepmeat industry

Abstract

A simulation study, based on a closed sheepmeat breeding nucleus and incorporating marker-assisted selection (MAS), was used to evaluate response when different proportions of animals were marker typed. Two traits were included in the simulation: trait 1, a production trait where phenotypes were available prior to selection, and trait 2, a carcass trait where phenotypic information was not available on breeding animals. Selection on an index which comprised estimated breeding values (EBVs) for both traits was possible by calculating EBVs for trait 2 as the sum of EBVs for a polygenic component, obtained from a genetic regression, and EBVs for a major gene component, obtained by inferring genotypes at a major gene locus from a linked marker locus. Different marker typing strategies were evaluated. These differed in the criteria used to select progeny for typing, and in the proportion of male and female progeny selected. Typing progeny that were likely to be used as a breeding animal, but excluding those whose marker genotype could be predicted with reasonable certainty, was an efficient genotyping strategy. Close to maximum gain at the major gene locus was achieved when only a proportion of animals were marker typed (for example 90% of maximum response was achieved with a little over one-third of the selection candidates marker typed). This indicate the potential for substantial savings in relation to the cost of marker typing in commercial breeding flocks.