Fig. 3

The prediction model for BSI. a Developed nomogram. The first and third lines represented the score and the total score scales, respectively; the last line represented the probability of BSI corresponding to the total score; and the peaks represented density. b Histogram of AUC frequency under 1000 sampling. c Calibration curves. AUC, area under the curve; CI, confidence interval. d The family scores’ boxplot in the BSI and non-BSI groups. e Prediction results’ confusion matrix. f Receiver operating characteristic (ROC) curves of family score and Shannon diversity at the family level. The student’s t-test was used for data with a homogeneous variance; otherwise, the Mann–Whitney test was used. Multiple groups were compared using one-way ANOVA