Figure 2
Figure 2. Sensitivity and specificity of aGvHD pattern. (A) Box-and-whisker plots of classification factor F obtained for classification of each individual HSCT sample (GvHD indicates graft-versus-host disease diagnosed according to Przepiorka et al26; control, HSCT sample showing no evidence of GvHD) in the training set and the blinded set (Table 1). The boxes depict the quartiles Q1 and Q3 of each distribution, and the statistical medians are shown as horizontal lines in the boxes. The whiskers indicate 3/2 times the interquartile range of Q1 and Q3. (B) ROC curve (bold line, AUC = 0.85) of the proteomics pattern diagnosis. Using the GvHD-specific polypeptide panel from Table 3, classification factor F is used as variable in ROC analysis in the 599 samples of the blinded set (Table 1). Ninety-five percent confidence intervals (95% CI) are indicated by thin lines.

Sensitivity and specificity of aGvHD pattern. (A) Box-and-whisker plots of classification factor F obtained for classification of each individual HSCT sample (GvHD indicates graft-versus-host disease diagnosed according to Przepiorka et al26 ; control, HSCT sample showing no evidence of GvHD) in the training set and the blinded set (Table 1). The boxes depict the quartiles Q1 and Q3 of each distribution, and the statistical medians are shown as horizontal lines in the boxes. The whiskers indicate 3/2 times the interquartile range of Q1 and Q3. (B) ROC curve (bold line, AUC = 0.85) of the proteomics pattern diagnosis. Using the GvHD-specific polypeptide panel from Table 3, classification factor F is used as variable in ROC analysis in the 599 samples of the blinded set (Table 1). Ninety-five percent confidence intervals (95% CI) are indicated by thin lines.

Close Modal

or Create an Account

Close Modal
Close Modal