Skip to main content

Table 3 Accuracy on simulated 454 reads.

From: The effect of sequencing errors on metagenomic gene prediction

 

GeneMark

MetaGene

MGA

Orphelia

ESTScan

Error rate

Sens.2

Spec.3

Sens.2

Spec.3

Sens.2

Spec.3

Sens.2

Spec.3

Sens.2

Spec.3

0

91.0 ± 3.6

93.8 ± 4.8

95.4 ± 2.8

92.8 ± 2.4

94.6 ± 2.7

94.1 ± 2.5

88.4 ± 3.5

96.7 ± 1.7

81.3 ± 7.8

97.9 ± 1.4

0.0022

85.3 ± 4.2

90.4 ± 5.6

89.3 ± 3.1

89.2 ± 3.5

89.6 ± 3.3

90.8 ± 2.6

80.0 ± 4.2

94.7 ± 2.1

77.2 ± 9.0

97.2 ± 1.5

0.0049

79.5 ± 4.9

87.6 ± 6.4

83.7 ± 3.5

85.9 ± 3.9

84.7 ± 4.0

87.7 ± 2.8

70.9 ± 5.9

92.5 ± 2.1

71.7 ± 11.5

96.2 ± 1.7

0.028

36.8 ± 4.9

68.3 ± 8.0

39.6 ± 3.9

60.6 ± 8.8

43.3 ± 5.5

61.9 ± 3.6

26.3 ± 9.1

68.3 ± 5.0

26.4 ± 11.2

86.2 ± 4.7

  1. The gene prediction accuracy (mean and standard deviation over all species in the simulated metagenome) of four metagenomic gene prediction tools GeneMark, MetaGene, MetaGeneAnnotator (MGA) and Orphelia, and of the EST processing tool ESTScan on simulated 454 reads is shown.