Skip to main content

Table 3 Comparison of various error correction software

From: Pollux: platform independent error correction of single and mixed genomes

  Illumina MiSeq - E. coli
  Errors Errors Reads Run
Software corrected (%) introduced (%) removed (%) time (min)
Pollux 93.81 1.28 0.89 3.01
Quake 58.78 0.05 0.92 4.09
SGA 78.65 0.09 1.11 15.51
BLESS 83.21 0.10 0.00 0.86
Musket 81.75 0.15 0.00 2.37
RACER 86.59 1.60 0.00 1.07
  Illumina genome analyzer II - S. aureus
  Errors Errors Reads Run
Software corrected (%) introduced (%) removed (%) time (min)
Pollux 87.04 0.38 31.73 3.67
Quake 75.30 0.10 29.81 4.81
SGA 47.45 0.02 10.71 14.28
BLESS 55.32 0.06 0.00 0.49
Musket 45.04 0.14 0.00 6.96
RACER 75.76 0.28 0.00 0.68
  Illumina HiSeq - L. pneumophila
  Errors Errors Reads Run
Software corrected (%) introduced (%) removed (%) time (min)
Pollux 96.16 0.13 6.01 14.40
Quake 99.66 0.00 4.25 22.25
SGA 84.61 0.03 3.87 73.82
BLESS 87.61 0.03 0.00 3.56
Musket 83.33 0.10 0.00 18.67
RACER 94.09 0.16 0.00 4.14
  Illumina HiSeq - M. tuberculosis
  Errors Errors Reads Run
Software corrected (%) introduced (%) removed (%) time (min)
Pollux 69.98 0.83 6.51 4.24
Quake 68.06 0.12 3.06 5.62
SGA 31.99 0.16 0.46 17.75
BLESS 60.01 0.13 0.00 1.04
Musket 44.68 0.80 0.00 5.93
RACER 65.14 1.28 0.00 0.79
  Roche 454 GS Junior - E. coli
  Errors Errors Reads Run
Software corrected (%) introduced (%) removed (%) time (min)
Pollux 81.02 4.14 0.82 5.45
Quake 0.21 0.00 0.07 3.62
SGA 8.94 3.63 1.64 5.33
BLESS 34.68 1.27 0.00 0.24
Musket 0.00 0.00 0.00 0.05
RACER 82.03 24.27 0.00 0.36
  Ion torrent PGM - E. coli
  Errors Errors Reads Run
Software corrected (%) introduced (%) removed (%) time (min)
Pollux 87.83 3.43 10.90 26.75
Quake 12.35 2.03 37.60 11.11
SGA 5.43 1.12 0.16 55.93
BLESS 22.82 0.52 0.00 1.25
Musket 9.40 4.88 0.00 47.27
RACER 67.86 15.95 0.00 1.64
  1. The evaluation is performed by aligning corresponding uncorrected reads and corrected reads, which were not removed, against a reference genome using SMALT. Corrected errors are an aggregate of all alignment errors which are found in uncorrected reads but not in corrected reads. Similarly, introduced errors are an aggregate of all alignment errors found in corrected reads but not in uncorrected reads and are relative to the sum of corrected and uncorrected errors. The percentage of reads removed by each software is noted. We note that Quake, SGA, and Musket were intended to only correct Illumina sequencing data.