Table 2.

Detailed information of the receiver operating characteristic (ROC) analysis for the constructive models and candidate lectins in the training cohort and validation cohort.

	AUC	95% confidence interval	Sig.	Cutoff value	Sensitivity	Specificity
Training cohort
Diagnosis of different types of LC
Model LC	0.926	0.886–0.965	<0.001	0.663	0.837	0.875
ECA	0.859	0.787–0.931	<0.001	0.018	0.894	0.775
RCA120	0.881	0.830–0.931	<0.001	0.021	0.772	0.875
HHL	0.870	0.809–0.931	<0.001	0.023	0.878	0.750
MAL-II	0.866	0.798–0.934	<0.001	0.025	0.837	0.825
PHA-E+L	0.810	0.737–0.834	<0.001	0.061	0.683	0.825
Model ADC	0.776	0.690–0.862	<0.001	0.544	0.787	0.729
DBA	0.634	0.532–0.735	0.013	0.052	0.520	0.729
PHA-E	0.634	0.529–0.738	0.013	0.016	0.720	0.583
Model SCC	0.649	0.548–0.749	0.006	0.614	0.641	0.644
Model SCLC	0.846	0.765–0.926	<0.001	0.804	0.710	0.867
STL	0.652	0.552–0.752	0.013	0.019	0.591	0.700
BS-I	0.618	0.501–0.735	0.052	0.018	0.785	0.467
Diagnosis of different stages of LC
Model LC	0.962	0.935–0.989	<0.001	0.666	0.908	0.875
ECA	0.893	0.827–0.959	<0.001	0.018	0.949	0.775
RCA120	0.915	0.833–0.951	<0.001	0.019	0.735	1.000
HHL	0.892	0.870–0.960	<0.001	0.023	0.908	0.750
Model LC-ES	0.941	0.899–0.983	<0.001	0.753	0.818	0.937
PTL-II	0.702	0.587–0.816	0.001	0.025	0.939	0.406
PWM	0.791	0.694–0.888	<0.001	0.017	0.697	0.750
LCA	0.708	0.593–0.823	0.001	0.016	0.742	0.687
WGA	0.709	0.599–0.819	0.001	0.031	0.909	0.406
Validation cohort
Diagnosis of different types of LC
Model LC	0.961	0.929–0.994	<0.001	0.754	0.918	0.939
Model ADC	0.619	0.488–0.750	0.079	0.569	0.706	0.586
Model SCC	0.693	0.567–0.820	0.004	0.578	0.800	0.667
Model SCLC	0.718	0.585–0.851	0.004	0.728	0.721	0.684
Diagnosis of different stages of LC
Model LC	0.935	0.885–0.984	<0.001	0.658	0.906	0.879
Model LC-ES	0.856	0.762–0.950	<0.001	0.668	0.829	0.810