Discovery of dominant and dormant genes from expression data using a novel generalization of SNR for multi-class problems

Table 4 Details of the selected genes by the frequency-based method for the Lung Cancer data set

	Class	Probe ID	Gene Symbol	Frequency	Ave. GDI	p-value	q-value
Dom	Adeno	38261_at	ABCC3	100	0.84	2.68E-05	0.0005
		35276_at	CLDN4	77	0.66	0.0004	0.0038
		39339_at	TMEM63A	63	0.64	0.0005	0.0044
		1930_at	ABCC3	43	0.68	0.0006	0.0051
		1802_s_at	ERBB2	23	0.68	0.0053	0.0280
	Normal	36119_at	CAV1	99	2.13	0	0
		1815_g_at	ESTs	36	2.14	0	0
		40994_at	GRK5	35	2.00	0	0
		36569_at	CLEC3B	34	1.83	0	0
		1814_at	ESTs	33	1.75	0	0
	SCLC	893_at	UBE2S	52	2.64	0	0
		39990_at	ISL1	47	2.38	0	0
		32272_at	TUBA1B	47	2.05	0	0
		894_g_at	UBE2S	43	2.18	0	0
		39605_at	FOXG1	38	2.78	0	0
	SQ	613_at	KRT5	100	1.48	0	0
		31791_at	TP63	96	1.26	0	0
		1898_at	TRIM29	74	1.18	3.19E-07	1.61E-05
		39016_r_at	KRT6A	56	1.18	3.19E-07	1.51E-05
		41266_at	ITGA6	38	1.21	9.57E-07	4.21E-05
	COID	40825_at	MAPRE3	67	2.47	0	0
		32254_at	VAMP2	52	2.31	0	0
		40165_at	TSPYL2	49	2.36	0	0
		41107_at	SNPH	39	2.80	0	0
		198_g_at	NME3	26	2.31	0	0
Dor	Adeno	36209_at	BRD2	97	0.68	0.0008	0.0176
		39799_at	FABP5	74	0.66	0.0022	0.0305
		40580_r_at	PTMS	67	0.58	0.0037	0.0399
		1315_at	OAZ1	57	0.61	0.0038	0.0408
		39561_at	CBX6	39	0.59	0.0050	0.0474
	Normal	36133_at	DSP	55	1.73	0	0
		31850_at	GCLC	41	1.58	0	0
		1248_at	POLR2H	23	1.48	0	0
		35194_at	GPX2	23	1.46	0	0
		39353_at	HSPE1	20	2.29	6.38E-07	0.0003
	SCLC	33908_at	CAPN1	54	1.30	6.38E-07	0.0003
		1109_s_at	PDGFA	36	1.71	1.50E-05	0.0022
		36952_at	HADHA	27	1.28	1.26E-05	0.0019
		338_at	ATF6	25	1.32	5.70E-05	0.0046
		36890_at	PPL	20	1.32	4.55E-05	0.0042
	SQ	38113_at	SYNE1	34	1.51	4.90E-05	0.0043
		33118_at	SEMA3B	31	1.36	3.99E-06	0.0009
		40665_at	FMO3	26	1.48	0.0003	0.0115
		37908_at	GNG11	19	1.52	2.11E-05	0.0024
		33267_at	ATP8A1	19	1.22	8.14E-06	0.0015
	COID	33322_i_at	SFN	79	2.67	0	0
		39728_at	ESTs	39	1.75	0	0
		36879_at	ECGF1	35	2.24	0	0
		925_at	ESTs	25	1.64	0	0
		33143_s_at	SLC16A3	24	2.07	4.79E-07	0.0003

Using the training data, for each class the top five most frequently occurring genes are selected.

ISSN: 1471-2105