Table 2.

The performance of models in the test set for each gene

Method	Gene	AUC	TP/FN/TN/FP	Sensitivity (95% CI)*	Specificity (95% CI)a
Binary relevance	IDH	0.964 (0.922–1)	5/0/107/14	100 (47.8–100)	88.4 (81.3–93.5)
	ATRX	0.775 (0.645–0.905)	9/8/104/5	52.9 (27.8–77%)	95.4 (89.6–98.5)
	MGMT	0.653 (0.538–0.769)	15/17/73/21	46.9 (29.1–65.3)	77.7 (67.9–85.6)
	EGFR	0.753 (0.659–0.847)	18/26/82/0	40.9 (26.3–56.8)	100 (95.6–100)
Ensemble classifier chain	IDH	0.967 (0.921–1)	5/0/107/14	100 (47.8–100)	88.4 (81.3–93.5)
	ATRX	0.822 (0.687–0.957)	12/5/93/16	70.6 (44–89.7)	85.3 (77.3–91.4)
	MGMT	0.761 (0.669–0.853)	21/23/80/2	47.7 (32.5–63.3)	97.6 (91.5–99.7)
	EGFR	0.743 (0.642–0.844)	26/6/55/39	81.2 (63.6–92.8)	58.5 (47.9–68.6)

AUC area under the receiver operating characteristic curve, TP true positive, FN false negative, TN true negative, FP false positive, CI confidence interval

^aSensitivity, specificity, and their 95% confidence intervals are expressed as percentages