Fig. 10

ROC curve analysis for PDAC prediction using five Machine Learning methods across the taxonomic ranks phylum, family, and genus. At the phylum level, the logistic regression model (95% CI 0.74, 0.95) achieved the best AUC of 0.85. At the family level, the logistic regression model (95% CI 0.78, 0.97) reached the highest AUC of 0.88. At the genus level, the SVM model performed best, with an AUC of 0.87 (95% CI 0.78, 0.95). AUC, area under the receiver operating characteristic (ROC) curve