Table 2.

Comparison of Spearman's correlation coefficients between four different expression measures (RPKM, mseq, GPseq and POME) on transcripts with high over-dispersion and high spatial dependence in 12 real datasets

Sample	Na	RPKM	mseq	GPseq	POME
LnCaP_0	1495	0.6454	0.6232	0.6071	0.7362
LnCaP_24	1534	0.6248	0.5990	0.5828	0.7118
LnCaP_48	1518	0.6059	0.5963	0.5152	0.6795
VCaP_0	1525	0.6641	0.6452	0.6496	0.7106
VCaP_24	1523	0.6259	0.5783	0.6691	0.6841
VCaP_48	1500	0.6354	0.5094	0.6820	0.7022
aT34	1563	0.6705	0.6470	0.5214	0.6914
aT34N	1476	0.6322	0.6268	0.6203	0.6523
DU145F	1525	0.6254	0.6017	0.5326	0.6749
DU145F2	1529	0.6239	0.5984	0.5375	0.6809
VCaP	1689	0.6211	0.5534	0.4853	0.6301
RWPE	1548	0.6250	0.6079	0.5173	0.6730

In each dataset, the method with the highest Spearman's correlation coefficient is highlighted in bold.

^aThe number of transcripts measured in Illumina sequencing with high over-dispersion and high spatial dependence.