Automated ensemble assembly and validation of microbial genomes

BMC Bioinformatics

Table 1 GAGE-B assemblies versus iMetAMOS (iMA) assemblies on the R. sphaeroides dataset

Assembler	GAGE-B rank	GAGE-B N50	GAGE-B CN50	GAGE-B reference coverage	GAGE-B #short-range	GAGE-B #long-range	iMA rank	iMA N50	iMA CN50	iMA reference coverage	iMA #short-range	iMA #long-range
ABySS	5	21,441	21,307	99.64%	422	2	5	38,322	36,101	99.66%	674	0
MIRA	6	15,792	15,190	99.78%	424	15	4	52,034	46,977	99.84%	420	5
MaSuRCA	1	130,714	119,839	92.31%	532	2	1	163,762	139,231	92.95%	535	5
SGA	7	9,108	9,055	99.39%	295	0	7	3,657	3,655	91.96%	460	3
SOAPdenovo2	3	33,491	32,605	98.82%	343	1	3	87,036	65,337	99.86%	432	5
SPAdes	2	118,093	89,065	99.95%	605	1	2	118,214	81,505	99.76%	526	4
Velvet	4	23,979	23,230	99.10%	760	5	6	19,652	19,355	99.71%	1191	4

The table lists the relative ranking, assembly, and corrected N50 for each of the 7 assemblers common between GAGE-B and the iMetAMOS run using an automatically-selected k-mer. Assembly N50 was computed using the reference as the true genome size. Corrected N50 (CN50) was calculated as in GAGE [7]. The short-range errors are a sum of SNPs and short and long indels. The long-range errors are a sum of translocations, relocations, and inversions.

ISSN: 1471-2105