. 2013 Apr;20(4):359–371. doi: 10.1089/cmb.2012.0098

Table 2.

Comparison of Different Assemblers

Assembler^a	No. of contigs	N50	N75	Covered (%)^b	MA^c	MM^d	CG^e
EC215 dataset
Velvet	198	82776	42878	99.93	4	1.2	4223
SOAPdenovo	192	62512	35069	97.72	1	26.1	4141
IDBA	246	48825	25483	99.60	3	1.3	4170
SPAdes	385	86548	42441	99.53	2	3.7	4223
SPAdes single read	458	54858	30309	99.56	0	0.7	4239
Pathset	360	91829	56830	99.56	1	2.1	4249
EC500 dataset
Velvet	169	105637	57172	99.28	5	2.5	4095
SOAPdenovo	982	57167	31582	99.88	0	0.2	4196
IDBA	227	57827	34421	99.26	1	4.2	4158
SPAdes	215	95454	46490	99.80	4	1.7	4223
SPAdes single read	493	54666	35158	99.88	0	0.9	4215
Pathset	320	97971	58548	99.46	2	2.0	4252

^{^a}

For each column, the best assembler by each criteria is indicated in bold.

^{^b}

Percent of genome covered is the ratio of total number of aligned bases in the assembly to the genome size.

^{^c}

MA: Misassemblies are locations on an assembled contig where the left flanking sequence aligns over 1 kb away from the right flanking sequence on the reference.

^{^d}

MM: Mismatch (substitution) error rate per 100 kbp is measured in the correctly assembled contigs.

^{^e}

CG: Complete genes is the number of genes contained completely within assembly contigs (using E. coli gene annotations from www.ecogene.org).