77 Development of an algorithm to discriminate between plasmid- and chromosomal-mediated AmpC Table S2. Confusion matrix on performance of the 2-class decision tree model trained using the 84 E. coli strains of the training set on all sets Training set (n=84) % (CI) Reference Accuracy 0.90 (0.82-0.96) non-pampC pampC Sensitivity 0.81 (0.66-0.91) Prediction non-pampC 42 8 Specificity 1.00 (0.92-1.00) pampC 0 34 Total 42 42 Validation set 1 (n=47) % (CI) Reference Accuracy 0.91 (0.80-0.98) non-pampC pampC Sensitivity 0.86 (0.42-1.00) Prediction non-pampC 37 1 Specificity 0.93 (0.80-0.98) pampC 3 6 Total 40 7 Validation set 2 (n=41) % (CI) t Accuracy 0.85 (0.71-0.94) non-pampC pampC Sensitivity 0.85 (0.71-0.94) Prediction non-pampC 0 6 Specificity NA NA pampC 0 35 Total 0 41 CI, 95% confidence interval, NA: Not Applicable 4
RkJQdWJsaXNoZXIy MTk4NDMw