Skip to main content
. 2013 Aug 14;29(20):2579–2587. doi: 10.1093/bioinformatics/btt440

Table 1.

Performance comparison of residue contact prediction

Method Accuracy (%) Coverage (%) Accuracy (δ = 4) (%)
Comparison on the training data set
    TMHcon 25.9 3.5 78.5
    TMhhcp1 49.5 8.2 83.9
    TMhhcp2 45.8 7.4 83.8
    SVMcon 8.4 1.5 55.1
    SVMSEQ 13.0 2.8 60.9
    PSICOV 42.1 6.7 74.7
    MemBraina 61.4 10.1 89.1
    MemBrainb 62.0 10.2 90.4
Comparison on the independent data set
    TMHcon 23.6 3.0 83.4
    MEMPACK 36.2 10.4 63.0
    TMhhcp1 48.1 6.1 84.4
    TMhhcp2 48.6 6.1 81.8
    SVMcon 10.4 1.6 68.0
    SVMSEQ 17.7 2.4 66.0
    PSICOV 50.9 6.6 81.1
    MemBrain 64.1 8.3 89.6

Note: aResults obtained from 4-fold cross-validation.

bResults obtained from jackknife cross-validation.