Scores scatter plot of (A) PCA showing discrimination between endometriosis and controls, (B) PLS-DA showing improved discrimination between the two groups. (C) PLS-DA model cross-validation by permutation test statistics showing that the R2 (0.986) and Q2 (0.976) values of the original model are significantly greater than those of the 100 permutated models, indicating a good predictive ability. Variable extraction using (D) loading and (E) VIP plots showing that IL-10 is the most important variable for class separation.

Abbreviations: PCA, principal-component analysis; PLS-DA, partial least square discriminate analysis; VIP, variable importance plot; IL, interleukin; TNF, tumor necrosis factor; IFN, interferon; TGF, transforming growth factor; VEGF, vascular endothelial growth factor; MMP, matrix metalloproteinase; TIMP, tissue inhibitor of metalloproteinase; COX, cyclooxygenase.