Skip to main content
Fig. 4 | BMC Bioinformatics

Fig. 4

From: Hydropathicity-based prediction of pain-causing NaV1.7 variants

Fig. 4

Binary classification of missense SCN9A-gene mutation sites based on their median distance from HP’s boundary. a ROC-curve plot constructed from data of median distances between mutation sites and HP’s boundary (for construction of data set see Additional file 1: S9a). Optimal threshold value corresponds to specificity and sensitivity values of 0.791 and 0.805, respectively. Area under ROC curve is 0.787. b Visualization of ROC curve data. Optimal threshold value 18.13 Å is marked with black dashed line. Shaded area around median distance values indicates the 95\(\%\) confidence intervals. ROC curve is constructed in R [73] by using the pROC package [93]. Two sets of missense SCN9A-gene mutation sites are employed; a pain-related set containing IEM, PPD and SFN mutation sites, and a neutral set containing mutation sites which are not expected to associate with pain disease phenotypes (Additional file 1: S8)

Back to article page