Fig. 8: Multivariate regression analysis in women with PCOS and GK rats.

a Orthogonal partial least-squares discriminant analysis (OPLS-DA) score plot for women with PCOS (n = 45, red dots) and control women (n = 44, blue dots) based on seven reproductive criteria (Supplementary Fig. 2a and Supplementary Table 1). The OPLS-DA score plot comprised one orthogonal and one predictive components, R2X(cum) = 0.549, R2Y(cum) = 0.810, Q2(cum) = 0.765. b OPLS-DA score plots for GK (n = 16, red triangles) and Wistar (n = 17, blue triangles) rats based on 11 reproductive criteria (Supplementary Fig. 2b). The OPLS-DA score plot comprised three orthogonal and one predictive components, R2X(cum) = 0.697, R2Y(cum) = 0.962, Q2(cum) = 0.871. c, d Pearson’s correlation matrix between reproductive and metabolic parameters for women (c) and rats (d). The stronger the correlation, the darker the color of blue. e, f OPLS-DA score plot based on 15 metabolic and reproductive criteria (Supplementary Fig. 2c, d) for 3 months GK (n = 9, red triangles) and Wistar rats (n = 9, blue triangles) (e) and for 6 months GK (n = 6, red triangles) and Wistar rats (n = 8, blue triangles) (f). The OPLS-DA score plot comprised three orthogonal and one predictive components; at 3 months: R2X(cum) = 0.697, R2Y(cum) = 0.962, Q2(cum) = 0.871 and at 6 months. R2X(cum) = 0.439, R2Y(cum) = 0.950, Q2(cum) = 0.702. g, h S-plot obtained from the OPLS-DA regression at 3 months (g) and 6 months (h) to identify the contribution of the 15 variables in the model (Pareto scaled data set). n represents the number of biologically independent women or animals in each group. Source data are provided as a Source Data file.