Table 6

Specificity value for each method at the sensitivity = 50% level (128 proteins set)

Method

Specificity

Proportion of predicted residues

Union set (10)

87%

15%

Conseq

87%

14%

Intersection set (9)

77%

24%

Closeness

77%

23%

Connectivity (4)

65%

35%

2-connectivity (3)

73%

28%

Betweenness

74%

26%

1/cluster (5)

68%

33%

Hssp

87%

14%

1/eccentricity

68%

32%

Results are obtained in average over the set of 128 proteins. Numbers are obtained by calculating the average as in Table 1. For each method, we indicate the value of the specificity and the proportion of residues that are predicted when the sensitivity is equal to 50%.

Thibert et al. BMC Bioinformatics 2005 6:213   doi:10.1186/1471-2105-6-213