Table 2. ROC curve analysis of diagnostic indicators in differentiating GC from AG and SG.

	Regressionmodel	C16OH	C6DC	C6	C0	ARG
GCvsSG
AUC	0.9986	0.9779	0.9910	0.8793	0.7385	0.6818
Cut-off value		0.0555	0.227	0.0545	11.29	37.25
Sensitivity (%)	98.55	90.28	97.22	76.39	65.28	83.33
Specificity (%)	98.55	97.1	95.65	89.86	81.16	50.72
P value	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	0.0002
GCvsAG
AUC	0.9961	0.9765	0.9882	0.8789	0.7309	0.7229
Cut-off value		0.0555	0.237	0.0625	9.462	30.99
Sensitivity (%)	98.55	87.84	98.65	82.43	54.05	75.68
Specificity (%)	100	97.1	94.2	79.71	91.3	63.77
P value	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
GCvs(AG+SG)
AUC	0.9977	0.9772	0.9896	0.8791	0.7347	0.7026
Cut-off value		0.0555	0.237	0.0625	10.59	37.25
Sensitivity (%)	99.32	89.04	97.95	83.56	59.59	83.56
Specificity (%)	98.55	97.1	94.2	79.71	85.51	50.72
P value	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001