Skip to main content

Table 2 The performance of MEGAN, SOrt-ITEMS, and SeMeta on the simulated datasets at different taxonomic levels - The scenario of unknown species

From: A novel semi-supervised algorithm for the taxonomic assignment of metagenomic reads

Method

 

Genus

Family

Order

Class

  

level

level

level

level

Dataset ds1

     

MEGAN

S e n. A

50.35 %

51.63 %

59.38 %

60.15 %

 

P r e. A

70.36 %

72.14 %

82.98 %

84.05 %

SOrt-ITEMS

S e n. A

21.05 %

27.96 %

31.91 %

41.05 %

 

P r e. A

30.11 %

40 %

45.66 %

58.73 %

SeMeta

S e n. A

67.66 %

74.64 %

75.38 %

76.59 %

 

P r e. A

77.68 %

85.71 %

86.57 %

87.95 %

Dataset ds2

     

MEGAN

S e n. A

56.14 %

58.74 %

59.1 %

61.39 %

 

P r e. A

83.23 %

87.08 %

87.62 %

91.05 %

SOrt-ITEMS

S e n. A

31.54 %

39.69 %

40.05 %

52.6 %

 

P r e. A

48.05 %

60.46 %

61.02 %

80.14 %

SeMeta

S e n. A

78.45 %

78.53 %

78.55 %

83.06 %

 

P r e. A

86.05 %

86.13 %

86.15 %

91.09 %

Dataset ds3

     

MEGAN

S e n. A

32.18 %

52.48 %

56.32 %

61.78 %

 

P r e. A

41.43 %

67.57 %

72.51 %

79.54 %

SOrt-ITEMS

S e n. A

9.34 %

28.5 %

34.19 %

44.42 %

 

P r e. A

12.45 %

37.98 %

45.57 %

59.2 %

SeMeta

S e n. A

37.41 %

58.64 %

60.38 %

71.46 %

 

P r e. A

49.37 %

77.39 %

79.68 %

94.31 %

  1. The bold values indicate the best results among the algorithms in the aspect of s e n s i t i v i t y A (S e n. A ) or p r e c i s i o n A (P r e. A )