Quake: quality-aware detection and correction of sequencing errors

Table 3 Velvet E. coli assembly

	Contigs	N50	N90	Scaffolds	N50	N90	Breaks	Miscalls	Cov
Uncorrected	398	94,827	17,503	380	95,365	23,869	5	456	0.9990
Corrected	345	94,831	25,757	332	95,369	26,561	4	315	0.9992

Velvet assemblies of E. coli 36 bp paired end reads at 152× coverage. After correcting the reads, more reads are included in the assembly into fewer contigs and scaffolds. N50 and N90 values were computed using the genome size 4,639,675 bp. The N50 value was similar for both assemblies, but N90 grew significantly with corrected reads. Correcting the reads also improved the correctness of the assembly producing fewer mis-assembled scaffolds (Breaks) and miscalled bases (Miscalls) and covering a greater percentage of the reference genome (Cov).

ISSN: 1474-760X