Table 2 Error correction benchmarking results for simulated Oxford Nanopore reads of various polyploid genomes (ploidy = 2, 3, 4)

From: VeChat: correcting errors in long reads using variation graphs

Method

#Reads

Error rate (%)

Mismatch (%)

Indel (%)

Haplotype coverage (%)

N50 (bp)

NGA50 (bp)

#Misassemblies

Ploidy=2

        

VeChat

30920

0.022

0.007

0.014

99.9

13,095

40,612

5

CONSENT

32661

0.212

0.160

0.052

99.9

13,040

40,903

37

Racon

31840

0.346

0.234

0.112

99.3

13,039

40,725

120

Canu

25506

0.390

0.206

0.183

100.0

13,820

40,846

6

Daccord

31438

0.438

0.410

0.027

99.2

12,987

40,730

3

Ploidy=3

        

VeChat

45113

0.090

0.041

0.050

100.0

13,130

40,497

90

CONSENT

49826

0.298

0.221

0.077

99.9

13,037

40,877

103

Racon

48520

0.673

0.501

0.172

98.6

13,028

39,286

631

Canu

36962

0.748

0.453

0.295

99.9

14,141

40,605

59

Daccord

49033

0.821

0.751

0.069

99.7

12,712

39,193

11

Ploidy=4

        

VeChat

58739

0.169

0.098

0.071

99.7

13,129

40,450

177

CONSENT

65384

0.292

0.195

0.097

100.0

13,033

40,864

121

Racon

63670

0.666

0.469

0.197

94.9

13,012

40,007

668

Daccord

65072

0.840

0.784

0.056

96.9

12,606

39,076

22

  1. The average sequencing coverage per haplotype is 30× and sequencing error rate is 10%.