Skip to main content

Table 1 Prediction for 5' and 3' gene-ends for five programs on test sets. Comparison of prediction for 5' and 3' ends of genes are performed among MED 2.0 (MED), Glimmer 2.02 post-processed by RBSfinder (GL2), Glimmer 3.02 (GL3) GeneMarkS (GMK), ZCURVE 1.0 (ZCV) and EasyGene (EG) on a set of reliable test setsa

From: MED: a new non-supervised gene prediction algorithm for bacterial and archaeal genomes

Test setb

Gene #

3' end match (%)

Both ends match (%)

  

MED

GL2

GL3

GMKc

ZCV

EG

MED

GL2

GL3

GMKc

ZCV

EG

Bsub_All

4100

98.7

98.2

97.6

98.9 (96.7)

98.4

94.5

83.8

75.0

82.4

86.1 (83.2)

83.1

79.6

EcoGene

854

99.1

99.3

99.4

99.9 (-)

98.8

99.4

92.0

82.5

91.9

93.8 (-)

89.2

91.1

Link

195

99.0

100.0

100.0

100.0 (100.0)

100.0

100.0  

93.3

85.6

94.4

94.4 (94.4)

92.3

92.1

EcoGene_short

58

93.1

91.4

96.6

100.0 (-)

86.2

93.3

91.4

77.6

89.7

98.3 (-)

77.6

90.0

Bsub123

123

95.1

91.1

87.8

97.6 (91.9)

91.9

73.0

85.4

73.2

77.2

87.8 (82.9)

78.0

66.0

Bsub72

72

94.4

91.7

87.5

98.6 (94.4)

93.1

82.4

87.5

75.0

77.8

93.1 (88.9)

86.1

76.5

Bsub51

51

92.2

88.2

82.3

98.0 (94.1)

90.2

84.8

90.2

70.6

78.4

94.1(90.2)

84.3

81.8

Psaer107

107

97.2

100.0

95.3

93.5 (-)

95.3

100.0

93.5

83.2

90.6

85.0 (-)

91.6

88.0

Mtub66

66

95.5

98.5

97.0

98.5 (-)

97.0

97.5

87.9

60.6

80.3

80.3 (-)

75.8

82.5

SolfGene

56

100.0

100.0

100.0

100.0 (-)

100.0

100.0

89.3

50.0

87.5

85.7 (-)

73.2

89.3

  1. aPrograms MED, GL2 (post-processed by RBSfinder), GL3 and ZCV were run locally, while GMK was run online, as described in the text. Predictions for EG were downloaded from [34].
  2. bExperiment confirmed TISs data sets: the first three represent two well-studied genomes: B. subtilis (Bsub_All) and E. coli (EcoGene and Link); the fourth to seventh represent short genes for E. coli (EcoGene_short) and B. subtilis (Bsub123, Bsub72 and Bsub51); Psaer107 and Mtub66 are selected for two GC rich genomes, M. tuberculosis (GC%: 65.6) and P. aeruginosa (GC%: 66.6); SolfGene corresponds to the archaeal S. solfataricus.
  3. cNumbers in parentheses indicate that the results of GeneMarkS have been reported in literature, (-) means no data reported.