824
New Simulation Method to Create Data Sets With a Desired Genetic Trend

Tuesday, August 19, 2014
Posters (The Westin Bayshore)
Anna-Maria Tyrisevä , MTT Agrifood Research Finland, Jokioinen, Finland
Martin H. Lidauer , MTT Agrifood Research Finland, Jokioinen, Finland
Gert P. Aamand , Nordic Cattle Genetic Evaluation, Aarhus, Denmark
Esa A. Mäntysaari , MTT Agrifood Research Finland, Jokioinen, Finland
Abstract Text: The simulation method consists of two main steps. First, original vector of observations is replaced with a vector of yearly increasing values that are used to predict pseudo breeding values (BVs) carrying a desired genetic trend. The pseudo BVs and the pedigree information are then used to calculate Mendelian sampling (MS) terms for each animal. Finally, simulated data sets carrying genetic trend are generated utilizing the MS terms created in the first step. The method retains the original structure of the pedigree and any number of random and fixed effects can be fitted.  The method is shown to yield data sets expressing closely the targeted genetic trend. The method was used to study the effect of genomic pre-selection on BVs and MS terms. The results revealed a clear bias in estimated BVs and MS terms in bulls after genomic pre-selection.

Keywords: evaluation bias, simulation of genetic trend, genomic pre-selection