Classification of Non-Small Cell Lung Cancer Using Significance Analysis of Microarray-Gene Set Reduction Algorithm

Table 1

Performance of SAMGSR on NSCLC data for stage segmentations.

	Training set				Test set
	Error (%)	GBS	BCM	AUPR	Error (%)	GBS	BCM	AUPR
(A) Trained on the microarray data (GSE50081)
No IC filtering, on stage (115)	1.18	0.050	0.809	0.976	32	0.318	0.51	0.612
No IC filtering, for AC (83)	0	0.039	0.825	0.996	35.7	0.357	0.5	0.627
No IC filtering, for SCC (14)	7.14	0.082	0.758	0.957	43.6	0.308	0.511	0.513
With IC filtering, on stage (75)	5.92	0.067	0.784	0.964	36	0.344	0.56	0.535
With IC filtering, for AC (119)	0	0.043	0.810	0.996	42.9	0.350	0.609	0.630
With IC filtering, for SCC (26)	2.36	0.062	0.802	0.992	32.7	0.256	0.589	0.583
(B) Trained on the RNA-seq data
No IC filtering, on stage (52)	0	0.028	0.871	0.997	30.8	0.270	0.523	0.529
No IC filtering, for AC (14)	11.43	0.087	0.779	0.961	58.4	0.454	0.533	0.536
No IC filtering, for SCC (28)	0	0.035	0.842	0.991	45.2	0.278	0.532	0.563
With IC filtering, on stage (24)	12.8	0.110	0.725	0.873	38.6	0.272	0.569	0.623
With IC filtering, for AC (31)	0	0.033	0.848	0.995	30.7	0.258	0.533	0.576
With IC filtering, for SCC (10)	9.09	0.101	0.712	0.905	33.3	0.279	0.556	0.641

Note: the test set is RNA-seq data in part (A) and GSE50081 microarray data in part (B).