Fig. 6
From: p53, cathepsin D, Bcl-2 are joint prognostic indicators of breast cancer metastatic spreading
PLS-DA cross-validation and performance. a Bar plot showing the performance measures (R2Ycum and Q2cum) using different numbers of components. b R2X: portion of the variation of X explained by specified principal component; R2X (cum) Cumulative explained portion of X set variation; Eigenvalue: number of variables (K) times R2X; R2Y: portion of the Y set variation modeled by the principal component; R2Y (cum): cumulative modeled variation of Y set; Q2: overall cross-validated R2 for the specific principal component; Limit: threshold cross-validation for the specific principal component; Q2 (cum): cumulative Q2 up to the specified component, is a model predictive power according to cross validation. Unlike R2X (cum), Q2 (cum) is not additive. c Permutation tests for: metastatic (left) and non metastatic tumors (right). Permutation tests were performed by comparing R2and Q2 of the original model with R2 and Q2 of Y-class-permutated models. The correlation coefficients of original and permuted data are reported on the X axis; 200 random permutations were carried out. The values of R2 and Q2 are reported on the Y axis. The green triangles and blue squares in the upper right (ρ = 1) correspond to the values of R2 (green triangles) and Q2 (blue squares) of the original data. The low values of intercepts show that the model has high statistical significance (no over-fitting)