. 2011 May 13;12(5):3148–3161. doi: 10.3390/ijms12053148

Table 1.

Summary of the partial least square regression models using amino acid compositions (AA), groups of amino acids (gAA) and sums of 3-z amino acid scales (Σz_i) of the food protein hydrolysates. The multiple correlation coefficient (R²) estimates the model fit whereas the cross-validated correlation coefficient (Q²_cv) indicates the models’ predictive powers.

Antioxidant Property	N^a	X^b	A^c	R²^d	Q²_cv^e	Permutation Test^f
Antioxidant Property	N^a	X^b	A^c	R²^d	Q²_cv^e	Int.R²_cum	Int. Q²_cv
DPPH-scavenging	16	AA + gAA+ Σz_i	1	0.407	0.327	0.241	−0.096
		AA	1	0.448	0.358	0.288	−0.108
		gAA	1	0.405	0.231	0.149	−0.043
		Σz_i	1	0.304	0.288	0.006	−0.127
Ferric reducing	16	AA + gAA + Σz_i	1	0.647	0.604	0.166	−0.208
		AA	1	0.728	0.668	0.223	−0.219
		gAA	1	0.536	0.531	0.073	−0.169
H₂O₂-scavenging	11	AA + gAA+ Σz_i	1	0.400	0.137	0.261	−0.122
		AA	1	0.467	0.136	0.314	−0.109
		Σz_i	1	0.340	0.232	0.088	−0.008
O₂⁻-scavenging	16	AA + gAA + Σz_i	1	0.394	0.179	0.212	−0.063
		gAA	2	0.602	0.142	0.166	−0.223

N, number of observations used for PLS analysis;

X, X-variables (descriptors) in the validated PLS models;

A, number of significant components used in PLS modeling;

R², multiple correlation coefficients;

Q²_cv, cross-validation correlation coefficients;

Permutation test, R²_cum and Q²_cv intercepts were calculated by SIMCA-P software during model validation.