Skip to main content

Table 2 Initial assessments of the effects of tuple length and Markov order of the background sequences on the performance of MaxBin+ \( {d}_2^S\mathrm{Bin} \) in terms of recall, precision and ARI for dataset 10genome-80×

From: Improving contig binning of metagenomic data using \( {d}_2^S \) oligonucleotide frequency dissimilarity

10genome-80×

Recall(%)

Precision(%)

ARI(%)

MaxBin

93.48

93.48

90.96

MaxBin+\( {d}_2^S\mathrm{Bin} \)

k = 4

r = 0

96.42

96.42

95.57

r = 1

93.99

93.99

90.86

r = 2

86.35

86.35

76.18

k = 5

r = 0

96.83

96.83

96.03

r = 1

95.40

95.40

93.19

r = 2

92.53

92.53

87.72

r = 3

59.71

60.91

37.18

k = 6

r = 0

96.93

96.93

96.05

r = 1

96.01

96.01

94.57

r = 2

94.24

94.24

91.40

r = 3

81.28

83.77

71.56

k = 7

r = 0

94.41

94.41

92.08

r = 1

93.26

93.26

91.92

r = 2

92.42

92.42

90.67

r = 3

65.88

77.73

50.48

k = 8

r = 0

88.26

82.94

80.04

r = 1

87.17

88.09

84.78

r = 2

87.19

87.12

82.73

r = 3

60.08

73.08

46.46

  1. The optimal numbers with respect Markov order are in bold