Table 4.
Summary of the modeling information of OPLS analysis.
| Name | No | R 2 X cum | R 2 Y cum | Q 2 Y cum |
|---|---|---|---|---|
| Model 1 | 1P + 1O | 0.105 | 0.887 | 0.582 |
| Model 2 | 1P + 3O | 0.501 | 0.815 | 0.215 |
| Model 3 | 1P + 2O | 0.42 | 0.568 | 0.216 |
| Model 4 | 1P + 4O | 0.552 | 0.804 | 0.258 |
| Model 5 | 1P + 2O | 0.293 | 0.731 | 0.351 |
| Model 6 | 1P + 2O | 0.471 | 0.621 | 0.361 |
Model 1 was generated by EZ group compared with DZ group with TCM symptoms; Models 2–6 were generated by urine metabonomic data (Model 2: EZ group compared with DZ group; Model 3: CHB group compared with normal control; Model 4: comparison of EZ, DZ, and WZ; Model 5: EZ compared with WZ; Model 6: DZ compared with WZ.)
No represents the sum number of components. For instance, “1P + 1O” means one predictive component and two orthogonal components for establishing the OPLS model.
R 2 X cum and R 2 Y cum represent the cumulative sum of squares (SS) of all the X's and Y's explained by all extracted components.
Q 2 Y cum is an estimate of how well the model predicts the Y's.