Roast: a tool for reference-free optimization of supertranscriptome assemblies

BMC Bioinformatics

Table 3 Evaluation of ROAST using real RNA-seq datasets

Organism	Assembly	Completeness			Fragmentation		False chimera	Inversion/translocation	Overall quality
Organism	Assembly	No. of bases \(^\text{a}\)	Mean contig length \(^\text{b}\)	Reference coverage \(^\text{c}\)	BLAST analysis	Proportion of read pairs on different contigs \(^\text{b}\)	Chimeric contigs \(^\text{c}\)	Proportion of read pairs with incorrect orientation \(^{b*}\)	TransRate score \(^\text{a}\)
Human	Initial	46,096,587	1109	0.09	3898	0.038	2184	0.011	0.46
Human	Improved	49,024,824	1271	0.10	3709	0.029	2529	0.013	0.51
Mouse	Initial	29,591,694	1131	0.07	2970	0.086	851	0.004	0.24
Mouse	Improved	30,458,066	1216	0.08	2830	0.060	877	0.003	0.28
Chicken	Initial	54,635,372	1018	0.21	2330	0.033	1747	0.003	0.51
Chicken	Improved	57,983,502	1072	0.23	2380	0.019	1690	0.002	0.55
Rice	Initial	26,514,527	701	0.17	2083	0.048	1253	0.002	0.40
Rice	Improved	30,336,262	784	0.18	1790	0.039	1151	0.001	0.44
Arabidopsis	Initial	32,430,389	646	0.24	11,373	0.067	4697	0.011	0.16
Arabidopsis	Improved	30,336,262	675	0.27	10,030	0.055	3725	0.011	0.40

Superior values shown in bold face Values calculated using:^a TransRate, ^b Samtools, ^c rnaQUAST.
* F1F2 and R1R2 orientations

ISSN: 1471-2105