Fig. 5

The results of OPLS-DA. AÂ score scatter plot, green dots and blue dots are located on both sides of the y-axis respectively which means model reconstructed by OPLS-DA under comprehensive consideration of the correlation of various parameters can effectively distinguish two groups.; (B) 200 times permutation tests, this test aimed to prove whether model was overfitting. The results were consistently lower than the original R2Y and Q2Y values, with the intercepts on y-axis of R2Y and Q2Y under 0.3 and 0.05, respectively which suggested that the model fitted well without overfitting; (C) VIP plot, The variable importance for the projection(VIP) means the importance of parameters in identifying different groups, VIP value more than 1 indicates that the parameter is significant and the higher the VIP value, the greater the importance of the parameter in distinguishing between different groups. ; (D) ROC curve of multiparametric prediction model