. 2024 Nov 14;19(11):e0310886. doi: 10.1371/journal.pone.0310886

Table 2. Prediction accuracy in univariate and multivariate using the same trait in two field conditions (Cond.).

		UV-CV1	MV-CV1					MV-CV2
Traits	Cond.	UV-GBLUP	MV-GBLUP	BBR	RKHS	SpikeSlab	RF	MV-GBLUP	BRR	RKHS	SpikeSlab
GY	HNW	0.58 ± 0.02^c	0.57 ± 0.02^c	0.58 ± 0.02^c	0.57 ± 0.02^c	0.58 ± 0.02^c	0.57 ± 0.02^c	0.67 ± 0.02^ab	0.69 ± 0.02^a	0.70 ± 0.02^a	0.66 ± 0.02^b
	LNR	0.53 ± 0.02^c	0.52 ± 0.02^c	0.55 ± 0.02^c	0.54 ± 0.02^c	0.53 ± 0.02^c	0.55 ± 0.02^c	0.67 ± 0.02^a	0.67 ± 0.02^a	0.67 ± 0.02^a	0.62 ± 0.02^b
TKW	HNW	0.57 ± 0.03^cd	0.55 ± 0.02^d	0.59 ± 0.02^c	0.58 ± 0.02^cd	0.56 ± 0.02^cd	0.55 ± 0.03^cd	0.73 ± 0.03^b	0.82 ± 0.01^a	0.83 ± 0.01^a	0.72 ± 0.03^b
	LNR	0.52 ± 0.02^cd	0.47 ± 0.05^d	0.51 ± 0.03^cd	0.53 ± 0.02^c	0.49 ± 0.03^cd	0.49 ± 0.03^cd	0.69 ± 0.06^b	0.81 ± 0.01^a	0.81 ± 0.01^a	0.71 ± 0.02^b
TW	HNW	0.74 ± 0.01^de	0.72 ± 0.01^e	0.74 ± 0.01^de	0.75 ± 0.02^d	0.74 ± 0.01^de	0.73 ± 0.01^de	0.85 ± 0.02^c	0.89 ± 0.01^ab	0.90 ± 0.01^a	0.87 ± 0.01^bc
	LNR	0.73 ± 0.01^c	0.72 ± 0.01^c	0.75 ± 0.01^c	0.75 ± 0.01^c	0.74 ± 0.01^c	0.73 ± 0.02^c	0.84 ± 0.05^b	0.89 ± 0.01^a	0.90 ± 0.01^a	0.88 ± 0.01^a
GYD	HNW	0.51 ± 0.02^cd	0.47 ± 0.04^d	0.52 ± 0.02^bc	0.52 ± 0.02^bc	0.51 ± 0.02^cd	0.52 ± 0.03^bc	0.56 ± 0.03^ab	0.60 ± 0.02^a	0.60 ± 0.03^a	0.56 ± 0.03^ab
	LNR	0.51 ± 0.02^de	0.49 ± 0.02^e	0.51 ± 0.02^de	0.53 ± 0.02^cd	0.52 ± 0.02^cde	0.52 ± 0.02^cde	0.58 ± 0.04^ab	0.59 ± 0.02^ab	0.60 ± 0.02^a	0.56 ± 0.03^bc
NDVI	HNW	0.43 ± 0.03^c	0.43 ± 0.02^c	0.44 ± 0.02^bc	0.44 ± 0.03^bc	0.42 ± 0.02^c	0.44 ± 0.02^bc	0.43 ± 0.03^bc	0.47 ± 0.02^ab	0.50 ± 0.03^a	0.46 ± 0.02^abc
	LNR	0.34 ± 0.02^c	0.38 ± 0.03^abc	0.36 ± 0.02^bc	0.34 ± 0.03^c	0.33 ± 0.03^c	0.37 ± 0.03^abc	0.42 ± 0.03^a	0.40 ± 0.03^ab	0.41 ± 0.03^a	0.38 ± 0.02^abc
GPC	HNW	0.63 ± 0.02^c	0.63 ± 0.02^c	0.63 ± 0.02^c	0.63 ± 0.02^c	0.62 ± 0.02^c	0.56 ± 0.03^d	0.79 ± 0.01^b	0.83 ± 0.01^a	0.84 ± 0.01^a	0.79 ± 0.02^b
	LNR	0.54 ± 0.02^c	0.51 ± 0.02^c	0.52 ± 0.02^c	0.53 ± 0.03^c	0.50 ± 0.03^c	0.44 ± 0.03^d	0.77 ± 0.02^ab	0.79 ± 0.02^a	0.81 ± 0.01^a	0.75 ± 0.02^b
GPD	HNW	0.56 ± 0.03^cd	0.57 ± 0.02^cd	0.57 ± 0.03^cd	0.58 ± 0.02^c	0.55 ± 0.02^de	0.52 ± 0.03^e	0.78 ± 0.01^a	0.79 ± 0.01^a	0.79 ± 0.01^a	0.74 ± 0.02^b
	LNR	0.52 ± 0.02^c	0.52 ± 0.02^c	0.53 ± 0.03^c	0.52 ± 0.03^c	0.48 ± 0.03^d	0.45 ± 0.03^d	0.77 ± 0.02^a	0.77 ± 0.02^a	0.78 ± 0.01^a	0.72 ± 0.02^b
YI	HNW	0.44 ± 0.03^b	0.34 ± 0.06^c	0.45 ± 0.02^b	0.44 ± 0.03^b	0.44 ± 0.03^b	0.46 ± 0.03^b	0.44 ± 0.07^b	0.59 ± 0.02^a	0.60 ± 0.02^a	0.54 ± 0.02^a
	LNR	0.48 ± 0.02^b	0.30 ± 0.06^c	0.47 ± 0.03^b	0.47 ± 0.03^b	0.48 ± 0.02^b	0.48 ± 0.02^b	0.44 ± 0.08^b	0.61 ± 0.02^a	0.62 ± 0.02^a	0.57 ± 0.02^a
FLA	HNW	0.62 ± 0.01^c	0.60 ± 0.02^c	0.62 ± 0.02^c	0.63 ± 0.02^c	0.63 ± 0.02^c	0.61 ± 0.02^c	0.62 ± 0.02^c	0.71 ± 0.02^a	0.71 ± 0.02^a	0.67 ± 0.02^b
	LNR	0.53 ± 0.02^cd	0.51 ± 0.02^d	0.55 ± 0.02^c	0.54 ± 0.02^cd	0.53 ± 0.02^cd	0.53 ± 0.03^cd	0.53 ± 0.03^cd	0.65 ± 0.02^a	0.65 ± 0.02^a	0.60 ± 0.02^b
DTHD	HNW	0.70 ± 0.02^c	0.67 ± 0.02^cd	0.69 ± 0.02^cd	0.68 ± 0.02^cd	0.69 ± 0.01^cd	0.66 ± 0.02^d	0.75 ± 0.02^ab	0.77 ± 0.01^a	0.77 ± 0.01^a	0.73 ± 0.02^b
	LNR	0.56 ± 0.03^d	0.55 ± 0.04^d	0.57 ± 0.02^d	0.57 ± 0.02^cd	0.57 ± 0.03^cd	0.58 ± 0.02^cd	0.62 ± 0.05^bc	0.68 ± 0.02^a	0.68 ± 0.02^a	0.63 ± 0.02^b
DTA	HNW	0.69 ± 0.02^b	0.65 ± 0.05^c	0.69 ± 0.02^b	0.68 ± 0.01^bc	0.69 ± 0.02^b	0.66 ± 0.02^bc	0.75 ± 0.01^a	0.75 ± 0.01^a	0.76 ± 0.01^a	0.74 ± 0.01^a
	LNR	0.52 ± 0.03^c	0.51 ± 0.03^c	0.53 ± 0.02^c	0.53 ± 0.02^c	0.53 ± 0.02^c	0.53 ± 0.03^c	0.59 ± 0.03^b	0.63 ± 0.02^a	0.64 ± 0.02^a	0.59 ± 0.02^b
DTM	HNW	0.67 ± 0.02^c	0.67 ± 0.02^c	0.68 ± 0.02^bc	0.67 ± 0.03^c	0.67 ± 0.02^c	0.65 ± 0.02^c	0.72 ± 0.01^a	0.70 ± 0.02^ab	0.72 ± 0.01^a	0.70 ± 0.02^ab
	LNR	0.48 ± 0.03^d	0.50 ± 0.02^cd	0.49 ± 0.03^cd	0.48 ± 0.03^d	0.48 ± 0.03^d	0.51 ± 0.03^cd	0.55 ± 0.03^ab	0.56 ± 0.03^ab	0.58 ± 0.03^a	0.53 ± 0.02^bc

Prediction accuracy (mean ± standard error) was observed using one univariate model as reference (UV-GBLUP) and five multivariate models such as MV-GBLUP, BRR, RKHS, Spike and slab regression, and Random Forest (RF). The multivariate model was carried out in two cross-validation schemes (MV-CV1 and MV-CV2) except for the model RF which was performed in MV-CV1 only. Duncan’s test was carried on for all output within each single row of the table. GY, grain yield; TKW, thousand kernel weight; TW, test weight; GYD, grain yield deviation; NDVI, normalized difference vegetation index; GPC, grain protein content; GPD, grain protein deviation; YI, yellow index; FLA, flag leaf appearance, DTHD, days to heading; DTA, days to anthesis; DTM days to maturity; HNW, high nitrogen and well-watered; LNR, low nitrogen and under rainfed.