Breeding without breeding: selection using the genomic best linear unbiased predictor method (GBLUP)

摘要

We demonstrate, using height data from a clonal trial, how the genomic best linear unbiased predictor method (GBLUP) is ideal for determining future breeding potential in situations (either in plantations or wild stands) where high mortality due to biotic or abiotic factors has occurred. The method is effective because it does not require the development of structured pedigree or classical progeny testing, rather it uses DNA fingerprinting to determine the genealogical relationship among individuals. The resulting genetic network is known as the realized relationship matrix, which in turn is used in classical quantitative genetics analyses to determine the genetic worth of all fingerprinted individuals. Selection of desirable individuals among the surviving population is aimed at maximizing genetic diversity even when the original genetic source is unknown. This is accomplished by determining the number of founder genome equivalents which can be used to estimate the inbreeding effective population size. During the selection phase, genetic diversity can be balanced against genetic gain so that diversity is maximized while gain for any particular attribute is optimized.