Figure 2

Imputing from low to high density has a limited effect on the accuracy of genomic selection. Correlation between results of genomic selection based on true and predicted genotypic values applying genomic selection for the original 90 k (Total-90 k) and 9 k SNP data sets (Total-9 k) for the total 371 lines, as well as for imputing low to high density marker data applying map-dependent (Beagle, FImpute, and IMPUTE2) and map-independent (Random Forest) imputation algorithms, for reference population sizes 50, 100, 200, and 300 out of 371 lines.