Meta-analysis of prostate cancer gene expression data identifies a novel discriminatory signature enriched for glycosylating enzymes

Barfeld, Stefan J; East, Phil; Zuber, Verena; Mills, Ian G

doi:10.1186/s12920-014-0074-9

Table 9 Comparison of the performance of a 31-gene signature with ERG, AR and KLK3 in discriminating between benign tissue, localised prostate cancer and metastatic disease

From: Meta-analysis of prostate cancer gene expression data identifies a novel discriminatory signature enriched for glycosylating enzymes

	id	agilent_probe	gene	auc.benign.vs.local	auc.benign.vs.met	auc.local.vs.met
1	10308	A_23_P334538	KLK3	0.85625	0.578125	0.55
2	18945	A_24_P108401	KLK3	0.5875	0.8203125	0.8375
3	24972	A_24_P344510	KLK3	0.6875	0.8359375	0.975
4	9045	A_23_P301414	ERG	1	0.75	0.666666666666667
5	15382	A_23_P57323	ERG	0.76875	0.5703125	0.5875
6	31809	A_24_P922378	ERG	0.713333333333333	0.683333333333333	0.5125
7	1039	A_23_P113111	AR	0.51875	0.75	0.75
8	18899	A_24_P106297	AMACR	0.98125	0.7890625	0.8
9	5760	A_23_P1833	B3GAT1	0.80625	0.5703125	0.7375
10	31242	A_24_P914625	BEND4	0.9	0.6875	0.675
11	2222	A_23_P128304	BICD1	0.90625	0.776785714285714	0.542857142857143
12	31375	A_24_P916586	BICD1	0.666666666666667	0.573333333333333	0.6
13	38999	A_32_P472968	BICD1	0.8125	0.5625	0.825
14	31242	A_24_P914625	BEND4	0.9	0.6875	0.675
15	14293	A_23_P45786	COL9A2	0.84375	0.578125	0.725
16	13511	A_23_P419760	CRISP3
17	12893	A_23_P403466	DLX1	0.71875	0.9453125	0.8
18	34905	A_32_P142818	DLX1	0.955555555555555	1	0.571428571428571
19	18202	A_23_P93058	DNAH5	0.944444444444444	0.94	0.555555555555556
20	19295	A_24_P120462	DNAH5	0.683333333333333	0.625	0.625
21	24667	A_24_P333461	DNAH5	0.555555555555556	0.5	0.611111111111111
22	26067	A_24_P388786	DNAH5	0.95	0.785714285714286	0.657142857142857
23	7096	A_23_P213424	ENC1	0.725	0.9453125	0.8
24	29137	A_24_P69095	ENC1	0.74375	0.517857142857143	0.657142857142857
25	18033	A_23_P91081	EPCAM	0.9625	0.78125	0.575
26	9045	A_23_P301414	ERG	1	0.75	0.666666666666667
27	15382	A_23_P57323	ERG	0.76875	0.5703125	0.5875
28	31809	A_24_P922378	ERG	0.713333333333333	0.683333333333333	0.5125
29	13214	A_23_P412214	RAP1GAP2	0.727272727272727	0.563636363636364	0.666666666666667
30	31091	A_24_P911927	RAP1GAP2	0.6	0.6	0.555555555555556
31	18139	A_23_P9232	GCNT1	0.975	0.705357142857143	0.7
32	5181	A_23_P16523	GDF15	0.7625	0.59375	0.675
33	7901	A_23_P250444	GJB1	0.86875	0.6328125	0.7375
34	26494	A_24_P404840	GJB1	0.85625	0.642857142857143	0.714285714285714
35	11669	A_23_P370666	GLYATL1	0.73125	0.53125	0.75
36	7727	A_23_P2344	HOXC6	0.785714285714286	0.958333333333333	0.857142857142857
37	147	A_23_P101806	HPN	0.925	0.75	0.5125
38	13011	A_23_P406782	HPN	0.90625	0.703125	0.8
39	6231	A_23_P203933	ITPR2	0.84	0.761904761904762	0.942857142857143
40	9378	A_23_P310	MARCKSL1	0.93125	0.921875	0.6625
41	2744	A_23_P13442	MICAL2	0.716049382716049	0.62962962962963	0.518518518518518
42	7851	A_23_P24843	MICAL2	0.8	0.5625	0.75
43	7314	A_23_P215956	MYC	0.8125	0.7578125	0.9125
44	20716	A_24_P178011	MYC	0.626666666666667	0.733333333333333	0.857142857142857
45	25948	A_24_P38363	MYC	0.5	1	1
46	3126	A_23_P139327	OR51E2	0.85	0.645833333333333	0.683333333333333
47	22188	A_24_P235756	OR51E2	0.88125	0.5625	0.8375
48	8971	A_23_P29816	PLA1A	0.925	0.640625	0.725
49	23649	A_24_P294408	PLA1A	0.9625	0.580357142857143	0.885714285714286
50	15416	A_23_P5778	RAB17	0.928571428571429	0.830357142857143	0.675
51	7973	A_23_P251387	REPS2	0.9875	0.78125	0.8125
52	33496	A_32_P100109	REPS2	0.94375	0.7890625	0.9625
53	6876	A_23_P211110	SIM2	0.914285714285714	0.803571428571429	0.7375
54	9061	A_23_P301886	SIM2	0.8	0.916666666666667	0.5
55	15035	A_23_P52939	SLC43A1	0.925	0.671875	0.6625
56	22855	A_24_P260443	THBS4	0.8125	0.8046875	0.9375
57	1849	A_23_P123503	TRIB1	0.88125	0.642857142857143	0.528571428571429
58	22636	A_24_P252497	TRIB1	0.775	0.578125	0.825
59	4805	A_23_P160460	UAP1	0.9625	0.7265625	0.9625

Data were downloaded from Grasso et al.,. ROC statistics were computed in an evaluation sample set having established the weighting for genes in the signature using logistic regression in a test sample set. We report the area under the curve (AUC) for each transcript and for the signature for each of three pairwise comparisons as generated using the R package ROCR.

Back to article page

ISSN: 1755-8794

Contact us

Submission enquiries: bmcmedicalgenomics@biomedcentral.com
General enquiries: ORSupport@springernature.com

BMC Medical Genomics

Contact us