Selection Index and the Best Linear Unbiased Prediction

Abstract

Selection index defined in this chapter is a linear combination of phenotypic values of the same trait collected from multiple relatives of a candidate individual. The selection index in its typical form has the same number of relatives and deals with the same types of relatives across all candidates. Therefore, the same set of index weights are involved in indices for all candidates. The best linear unbiased prediction (BLUP) is a generalization of selection index where phenotypic values of all measured individuals are included in the index to predict the breeding value of a candidate, but the weights of the index can vary from one candidate to another. Matrix algebra is used to calculate the weights of a selection index. The MIXED procedure in SAS is used to calculate the BLUP values of candidates.