A function accounting for training set size and marker density to model the average accuracy of genomic prediction.

We observed a decreasing trend in the weighting factor w when reducing the number of SNPs but the extent of the decrease was limited, so that even with 10′000 SNPs a high percentage of the genetic variance (71.2% for milk yield and 75.3% for somatic cell score, both with model D2) is captured and not much is gained by applying a more than four-fold SNP density.

also in 47,180 other papers