Table 1 Maize data sets
Trait-Environment Combination+Mean CorrelationMean PMSE
BRRBIRRBAIBIR1BIR2BRRBIRRBAIBIR1BIR2
FFL-WW0.8180.8190.8280.8420.8470.2620.2600.2190.2010.196
FFL-SS0.7540.7550.7590.7620.7640.3420.3250.3250.3240.320
MFL-WW0.8220.8220.8290.8410.8500.2630.2630.2250.2030.198
MFL-SS0.7760.7770.7840.7820.7880.3180.3180.2990.2980.293
ASI-WW0.5820.5820.5800.5780.5740.6490.6480.6490.6510.655
ASI-SS0.6130.6140.6170.6140.6110.6530.6520.6470.6490.650
Average0.7270.7280.7330.7360.739
% increase−1.2%−1.1%−0.5%0%0.4%
GY-SS0.3200.3260.3050.3540.3600.8900.8830.9040.8660.862
GY-WW0.5570.5580.5550.5580.5580.6770.6750.6800.6740.675
GY-HI0.6340.6350.6640.6670.6740.5970.5950.5580.5540.546
GY-LOW0.4100.4120.4190.4240.4230.8630.8450.8470.8350.837
Average0.4800.4830.4860.5010.503
% increase−4.2%−3.7%−3.1%0%0.4%
GLS 10.2410.2380.2600.2820.2870.9300.9280.9200.9020.900
GLS 20.4190.4210.4140.4270.4260.8270.8250.8350.8190.819
GLS 30.5880.5890.5850.5860.5840.6250.6240.6290.6270.629
GLS 40.5220.5220.5290.5330.5280.7350.7330.7280.7210.729
GLS 50.3380.3410.3480.3600.3560.8190.8150.8070.8010.803
GLS 60.2570.2740.2410.2780.2760.9750.9640.9860.9590.960
GLS 70.4740.4750.4720.4800.4840.7610.7610.7660.7560.750
GLS 80.5950.5960.5930.5930.5910.6180.6170.6210.6200.622
GLS 90.5220.5220.5290.5330.5300.7340.7320.7280.7230.727
Average0.4420.4420.4410.4530.445
% increase−2.9%−2.4%−2.6%0%−0.2%
NCBL 10.6490.6490.6970.6930.6970.5920.5920.5210.5270.523
NCBL 20.4690.4730.5230.5260.5240.7310.7230.6730.6790.680
Average0.5590.5610.610.6100.6110.6560.6580.6460.6380.639
% increase−9.0%−8.7%0.2%0%0.7%
  • Mean predicted correlation, average, % increase correlation with respect to model BIR1, and mean PMSE of five models, BRR, BIRR, BAI, BIR1, and BIR2, for 50 random partitions for each of 21 trait-environment combinations. The largest correlations and smallest PMSE for each trait-environment combination are in boldface. PMSE, predictive mean squared error; BRR, Bayesian ridge regression; BIRR, Bayesian inverse ridge regression; BAI, Bayes A inverse; BIR1, Bayesian inverse regression model 1; BIR2, Bayesian inverse regression model 2; FFL, female flowering; WW, well-watered environment; SS, severe drought stress; MFL, male flowering; ASI, MFL to FFL interval; GY, grain yield; HI, optimum environment; LOW, stress environment; GLS, Cercospora zeae-maydis; NCLB, Exserohilum turcicum.