Skip to main content

Table 5 Keyword recovery scores for the Operon method alone and in combination with the Rosetta Stone (RS), Phylogenetic Profile (PP), and conserved Gene Neighbor (GN) methods

From: Inference of protein function and protein linkages in Mycobacterium tuberculosis based on prokaryotic genome organization: a combined computational approach

Method

Number of links between SWISS-PROT annotated proteins

Keyword recovery %

Maximum false positive fraction

Keyword recovery of random links (100 trials) %

Random links SD %

Keyword recovery signal to noise

0 bp Operon

308

51

0.25

5.6

1

9.1

25 bp Operon

642

49

0.28

5.6

0.7

8.8

50 bp Operon

818

47

0.31

5.7

0.6

8.2

75 bp Operon

912

44

0.36

5.5

0.5

8

100 bp Operon

1044

45

0.35

5.7

0.5

7.9

100 bp Operon and RS

88

67

0.05

5.5

1.8

12

100 bp Operon and GN

638

60

0.16

5.5

0.7

11

100 bp Operon and PP

268

62

0.11

5.8

0.9

11

100 bp Operon, PP and GN

242

61

0.11

5.6

1.2

11

100 bp Operon, PP and RS

62

74

0.03

5.6

2

13

100 bp Operon, RS and GN

78

65

0.05

6.3

1.9

10

All four methods

54

70

0.04

5.7

2.3

12

  1. Keyword recovery and maximum false positive fraction calculated as described in Table 2. Random links were established between the same number of random pairwise Swiss-Prot annotated genes as exist real links between Swiss-Prot annotated genes (mean and standard deviation of 100 random trials indicated).
  2. Signal to Noise calculated as: