Design Of Reference Population Affects Reliability
POLAND - Genomic selection opens up opportunities to make selections to bring about new traits that are either expensive or difficult to measure, for example regarding methane emissions.
29 March 2011
1 minute read
With regard to such traits, it is important to maximise the achieved reliability of genomic prediction using a reference population of a limited size. The Animal Breeding and Genomics Centre (ABGC) and the Poznan University of Life Sciences therefore joined forces to investigate the effect of the reference population design in terms of family structure on the reliability of genomic selection in a simulation study. Composing the most suitable reference population is a matter of optimisation as there is a trade-off between obtaining a low level of relationship between the animals in the reference population and high level of relationship to the animals in the population to be evaluated.
The reference populations were small and consisted of cows with only own performance records. Analyses showed that optimally-designed reference populations for use in genomic prediction or genome-wide association studies should have a relatively loose family structure. Therefore, the average familial relationship level between the animals included in a reference population should be low. This is in contrast with designs for linkage mapping studies that typically have strong family structures. The relationship between the reference population and an evaluated animal should be maximised in order to achieve reliable predictions.
TheCattleSite News Desk
