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 2 Intra- and cross-platform comparison of DEG lists generated from simulated microarray and RNA-Seq data.

	T-test	Ebayes	SAM	baySeq	DESeq	SAMseq	NOISeq
T-test	100.0%
Ebayes	5.1%	100.0%
SAM	4.8%	95.1%	100.0%
baySeq	3.0%	48.1%	49.6%	100.0%
DESeq	3.3%	48.7%	50.2%	75.7%	100.0%
SAMseq	3.4%	36.5%	37.0%	54.1%	51.9%	100.0%
NOISeq	3.3%	39.7%	39.7%	46.5%	50.4%	46.9%	100%

The simulations were carried out as described in Methods. The overlap percentages was calculated based on the percent of true positive DEGs (95% minimum fold change method: FC ≥ 2 & FDR < = 0.05) identified by each of the three microarray algorithms (T-test, SAM, eBayes) and four RNA-Seq algorithms (SAMseq, baySeq, DESeq, NOISeq).
¹The overlap rate was calculated based on true positive DEGs called by each method. The microarray and RNA-Seq cross-platform DEG overlap rates are shown in bold style.

ISSN: 1471-2105