From: Efficient strategies for leave-one-out cross validation for genomic best linear unbiased prediction
M1
M2
M3
M4
M5
Phenotypes
1
2
1.97
0
2.12
3
–0.62