Table 5 Wheat data set 1
IBCFMFMEMTME
No MarkersWith MarkersNo MarkersNo MarkersWith MarkersNo MarkersWith Markers
Env–TraitMeanSEMeanSEMeanSEMeanSEMeanSEMeanSEMeanSE
Bed2IR_HD0.8750.0110.3740.0230.8640.011−0.0050.030−0.0220.0300.8290.0080.8800.006
Bed2IR_GNDVI0.6640.0120.4110.0200.6600.012−0.0120.023−0.0120.0230.1540.0270.0400.022
Bed2IR_GY0.5570.0120.3530.0180.4950.014−0.0140.0240.0430.0220.1810.0240.0980.025
Bed2IR_PH0.6470.0110.3250.0230.6170.011−0.0220.0160.0260.0150.2350.0140.0310.015
Bed5IR_HD0.8730.0070.3900.0150.8770.007−0.0060.0280.0240.0280.8660.010.8930.008
Bed5IR_GNDVI0.5860.0110.2490.0220.5650.012−0.0080.0190.0050.0190.0010.033−0.0030.033
Bed5IR_GY0.0910.0320.0350.0210.0770.0310.0190.0240.0240.0240.4030.0220.3930.027
Bed5IR_PH0.4100.0220.3450.0180.3830.0230.0630.021−0.0130.0250.6030.0120.5050.015
Drip_HD0.9200.0030.4640.0230.9090.0040.0180.0230.0060.0230.9170.0040.9220.003
Drip_GNDVI0.5680.0160.3710.0310.5630.0170.0470.034−0.0050.036−0.120.027−0.0430.026
Drip_GY0.4010.0240.3620.0180.4620.020−0.0250.027−0.0150.0270.4320.0190.3640.017
Drip_PH0.6570.0160.3350.0220.6600.0160.0030.017−0.0210.0170.5790.0220.5380.022
Average0.6040.0150.3350.0210.5940.0150.0050.0240.0030.0240.4240.0180.3850.018
  • Prediction accuracies with Pearson correlation for each environment–trait (Env–Trait) combination of the proposed methods for the wheat data set from Rutkoski et al. (2016), under cross-validation scheme CV1. The best predictions of the four methods are in boldface, and the comparisons are made by row. “No markers” means that genomic information was not used, while “with markers” means that genomic information was used. MF, matrix factorization.