Parallel comparison of Illumina RNA-Seq and Affymetrix microarray platforms on transcriptomic profiles generated from 5-aza-deoxy-cytidine treated HT-29 colon cancer cells and simulated datasets

Table 1 Cross-platform overlap in DEG lists using RNA-Seq and microarray HT-29 data.

Comparison (5-Aza)	Methods	Cuffdiff	SAMSeq	baySeq	DESeq	Noiseq
5 µM vs 0 µM	T-test	25.8%	22.5%	24.5%	25.8%	29.7%
	SAM	39.9%	39.9%	42.7%	44.5%	33.5%
	Ebayes	39.5%	39.5%	42.2%	44.0%	33.3%
10 µM vs 5 µM	T-test	0.5%	0.5%	0.6%	0.6%	0.0%
	SAM	31.1%	30.2%	32.6%	34.2%	19.4%
	Ebayes	30.3%	28.3%	31.9%	33.5%	19.1%

The cross-platform overlap rates (%)¹ were calculated between the three microarray DEG algorithms (T-test, SAM, eBayes) and the five RNA-Seq algorithms (Cuffdiff, SAMSeq, baySeq, DESeq, and NOISeq) as described in Methods. Comparisons were made to 1) measure the effect of treating HT-29 cells with 5 μM 5-Aza compared to control cells and 2) the effect of increasing the 5-Aza concentration from 5 μM to 10 μM.
¹The DEG overlap rates were calculated by dividing the number of overlapped DEGs over the union number of DEGs identified from both methods. Overlap rates higher than 40% are denoted in bold style.

ISSN: 1471-2105