Differentially expressed genes from RNA-Seq and functional enrichment results are affected by the choice of single-end versus paired-end reads and stranded versus non-stranded protocols

Table 2 Number of genes with significantly different counts

Experiment	1	2	3	4
Genes tested	15456	12702	14016	14166
biotype
protein coding	14328 (92.7%)	11891 (93.6%)	12009 (85.7%)	12530 (88.5%)
non protein coding	1027 (6.6%)	733 (5.8%)	1988 (14.2%)	1620 (11.4%)
- pseudogenes	232	145	452	396
- antisense	242	124	604	532
SEvsPE DEGs^a	111 (0.7%)	176 (1.4%)	327 (2.3%)	608 (4.3%)
biotype
protein coding^b	29 (26.1%)	93 (52.8%)	139 (42.5%)	396 (65.1%)
non protein coding^c	82 (73.9%)	74 (42%)	186 (56.9%)	208 (34.2%)
- pseudogenes^d	76 (68.5)	50 (28.4%)	105 (32.1%)	150 (24.7%)
- antisense^e	1 (0.9%)	3 (1.7%)	13 (4.0%)	12 (2.0%)
NSvsPE DEGs^f	NA	729 (5.74%)	1719 (12.26%)	1615 (11.4%)
biotype
protein coding^b		555 (76.1%)	1167 (67.9%)	1291 (79.9%)
non protein coding^c		167 (22.9%)	543 (31.6%)	317 (19.6%)
- pseudogenes^d		15 (2.1%)	79 (4.6%)	45 (2.8%)
- antisense^e		48 (6.6%)	271 (15.8%)	167 (10.3%)

^a Differential expression (DE) test of Controls in PE data (n = 3) versus Controls in SE data (n = 3) using edgeR, FDR < 0.05
^b Number of DE protein coding genes and proportion as a percentage of the tested genes
^c Number of DE non-protein-coding genes and proportion as a percentage of the tested genes
^dNumber of DE pseudogenes
^e Number of DE genes
^f DE test of Controls in PE data (n = 3) versus Controls in NS data (n = 3) using edgeR, FDR < 0.05

ISSN: 1471-2164