Table 2.

The 5 critical DEGs and the classifiers in the 7 datasets.

Dataset	Data source	Tumor	Non-tumor	Classifier	Intercept	CCDC107	CXCL12	GIGYF1	GMNN	IFFO1	Accuracy %	Sensitivity %	Specificity %
1	TCGA	374	50	CF1	−11.185			6.992	2.072	−7.164	68.87	64.71	100
				CF2	−10.186	3.028		−6.986		10.801	58.25	52.67	100
				CF3	3.311		−2.582	1.925	1.923		97.88	97.59	100
				CFmax							100	100	100
	Stage 1	173	50	CFi	−6.536				2.954		96.86	97.69	94
	Stage 1	173	50	CFi	4.707		−1.036				91.48	94.8	80
2	GSE54236	81	80	CF1	21.724	4.879	−2.086		9.677		78.88	61.73	96.25
				CF2	15.893		−5.440	0.355		14.239	78.26	64.2	95
				CFmax							90.06	85.19	95
3	GSE6764	35	10	CF1	13.3	5.098		2.756		5.33	92	92.31	90
				CF2	8.3	5.336			21.878	0.454	72	67.69	100
				CFmax							100	100	100
		35	13 CI	CF1	0.983	4.182	−12.002		−6.002		95.83	94.29	100
				CF2	10.090	3.002			2.502	10.154	89.58	85.71	100
				CFmax							100	100	100
		17 DN	10	CF1	9	8.5	−18.928		26.26		92.59	88.24	100
				CF2	2.2	12.686	−44.914	−0.974			81.48	70.59	100
				CFmax							100	100	100
4	GSE41804	20	20	CF1	1.936		−4.931	−7.614		11.438	92.5	90	95
				CF2	−49.690	6.308		−3.712	0.894		85	75	95
				CFmax							97.25	100	95
5	GSE25097	268	249	CF1	−1.697		−0.531	−3.858		13.330	83.75	70.15	98.39
				CF2	−2.834		−5.803		10.392	−6.829	96.13	92.54	100
				CFmax							98.65	98.88	98.39
6	GSE63898	228	168	CF1	−6.129	7.782	−8.426		3.364		97.47	96.49	98.81
	Stage 1	19	168	CFi	−11.295				1.202		95.19	63.16	98.81
	Stage 1	19	168	CFi	14.587		−1.719				97.33	84.21	98.81
7	GSE101685	24	8	CF1	−1.377		−5.030		6.625		100	100	100
8	GSE148355	54	15	CF1	3.557		−0.865		5.825		100	100	100
9	GSE164760	53	6	CF1	−11.425		−3.300		2.306	2.815	49.15	43.40	100
				CF2	−3.909	−4.797			9.897	−2.595	84.75	83.02	100
				CFmax							94.92	94.34	100
10	Our RNA-seq data	17	17	CF1	6.432	6.587	−2.235	5.964		−14.586	97.06	100	94.12
Total		1184	672								97.79	97.47	98.27

Abbreviations: DN, dysplastic nodule; CI, cirrhosis.The final classifiers are combined classifiers of individual competing factors.