Removal of redundant contigs from de novo RNA-Seq assemblies via homology search improves accurate detection of differentially expressed genes

Table 4 Summary of co-identity in redundancy-removed contigs

	Raw Contigs	Longest Contigs	Clustered Contigs	Annotated Contigs	Plant DB Contigs	PlantClust50 DB Contigs
Correlation coefficient of log2-fold change between gene dataset and reference sequences	0.60	0.60	0.88	0.88	0.83	0.73
No. of DECs exhibiting co-identity with DEGs^a	8602 (73.8 %)	8331 (74.2 %)	8506 (75.8 %)	8303 (74.0 %)	7594 (67.7 %)	5878 (52.4 %)
No. of DECs redundant or not exhibiting co-identity with DEGs^b	15,706 (64.6 %)	14,219 (63.1 %)	28,640 (77.1 %)	1372 (14.2 %)	9052 (54.4 %)	4809(45.0 %)
P-value^c	0.44	0.20	0.01	1.00	0.93	0.93
Fisher’s exact test^d	30	30	31	4	29	17

^a Number of contigs identical to DEGs in gene dataset (values in parentheses are percentage of all DEGs identical to DECs)
^b Number of contigs not corresponding DEGs in gene dataset (values in parentheses are percentage of contigs without corresponding genes in contig group)
^c Correlation between gene database and contig group in GO slim term distribution calculated by Kolmogorov–Smirnov test
^d No. of GO terms significantly different from gene dataset in annotation count

ISSN: 1471-2164