Table 5. ROC curve analysis of detected markers.
Comparison | Marker | AUC | P a | Best value of separation | Sensitivity % | Specificity % |
CP vs CTRL | ||||||
2-butanone | .756 | .027 | 2.90 | 75 | 79 | |
2- or 3-pentanone | .768 | .021 | 1.08 | 75 | 64 | |
C8-ketone | .815 | .006 | .10 | 83 | 64 | |
C9-ketone | .756 | .027 | ||||
Monoterpene | .887 | .001 | 2.16 | 83 | 86 | |
Terpene related | .810 | .007 | .39 | 83 | 64 | |
S-compound | .208 | .012 | .11 | 83 | 72 | |
Sulfoxide-compound | .756 | .027 | ||||
N-compound | .768 | .021 | .19 | 83 | 64 | |
Heptadienol | .768 | .021 | 1.48 | 83 | 72 | |
Methanol | .738 | .040 | 485.73 | 58 | 86 | |
CP-A vs CTRL | ||||||
C9-ketone | .786 | .048 | .099 | 67 | 99.7 | |
Monoterpene | .774 | .058 | 2.16 | 66 | 96 | |
N-compound | .929 | .003 | .26 | 83 | 94 | |
Dimethyl sulphide | .750 | .083 | 6.28 | 83 | 64 | |
CP-A vs CP-B+C | ||||||
2-butanone | .139 | .037 | 178.5 | 83 | 99 | |
2- or 3-pentanone | .167 | .055 | 1.13 | 1 | 67 | |
C8-ketone | .056 | .010 | .11 | 1 | 67 | |
Monoterpene | .000 | .004 | 6.7 | 1 | 99 | |
S-compound | .861 | .037 | 0.04 | 83 | 99 | |
NS-compound | .028 | .006 | .85 | 1 | 83 | |
N-compound | .917 | .016 | .14 | 83 | 99 | |
Heptadienol | .000 | .004 | 2.30 | 1 | 83 |
Null hypothesis: true area = 0.5.