Table 1. Parameter values of OPLS-DA models.
OPLS-DA model | Noa | R2X(cum)b | R2Xc | Q2(cum)d |
---|---|---|---|---|
Control vs. male | 1P + 2O | 0.537 | 0.213 | 0.785 |
Control vs. female | 1P + 2O | 0.631 | 0.292 | 0.823 |
Male vs. female | 1P + 2O | 0.874 | 0.465 | 0.87 |
Notes.
No, the number of predictive (P) and orthogonal (O) components.
The R2X(cum) value is the predictive and orthogonal variation in model samples (X) explained by the model.
The R2X value is the amount of variation in X which is correlated to Y (response matrix).
The Q2(cum) value describes the predictive ability of the model, based on sevenfold-cross validation.