Table 1.
Overall performance of ElliPro in comparison with other methods#.
ElliPro (radius 6Å) | PPI-PRED | PatchDock | ClusPro (DOT) | |||||||||
Best prediction (score 0.5) | Average (score 0.5) | 1st prediction (score 0.7) | ProMate (patch) | 1st patch | best patch | 1st model | best model of top 10 | 1st model | best model of top 10 | CEP (average) | DiscoTope (-7.7) | |
Overall statistics | ||||||||||||
sensitivity | 0.601 | 0.165 | 0.093 | 0.091 | 0.153 | 0.331 | 0.300 | 0.425 | 0.258 | 0.453 | 0.310 | 0.416 |
1-specificity | 0.138 | 0.109 | 0.047 | 0.083 | 0.161 | 0.135 | 0.135 | 0.114 | 0.079 | 0.067 | 0.223 | 0.214 |
precision | 0.291 | 0.119 | 0.158 | 0.101 | 0.083 | 0.188 | 0.175 | 0.262 | 0.235 | 0.390 | 0.110 | 0.155 |
accuracy | 0.840 | 0.832 | 0.879 | 0.841 | 0.780 | 0.819 | 0.816 | 0.846 | 0.863 | 0.892 | 0.739 | 0.754 |
AUC | 0.732 | 0.528 | 0.523 | 0.504 | 0.496 | 0.598 | 0.583 | 0.656 | 0.589 | 0.693 | 0.544 | 0.601 |
P-value | 0.00E+00 | 1.37E-04 | 9.52E-06 | 2.74E-01 | 1.0E+00 | 7.8E-30 | 9.0E-23 | 0.0E+00 | 7.9E-34 | 0.0E+00 | 4.3E-06 | 4.1E-25 |
Statistics averaged over epitopes | ||||||||||||
sensitivity | 0.58 ± 0.25 | 0.17 ± 0.10 | 0.10 ± 0.20 | 0.09 ± 0.17 | 0.15 ± 0.24 | 0.34 ± 0.32 | 0.29 ± 0.26 | 0.42 ± 0.29 | 0.25 ± 0.31 | 0.46 ± 0.28 | 0.34 ± 0.28 | 0.43 ± 0.31 |
1-specificity | 0.12 ± 0.10 | 0.11 ± 0.04 | 0.05 ± 0.03 | 0.08 ± 0.03 | 0.16 ± 0.07 | 0.14 ± 0.07 | 0.15 ± 0.06 | 0.13 ± 0.07 | 0.10 ± 0.07 | 0.08 ± 0.05 | 0.28 ± 0.20 | 0.22 ± 0.15 |
precision | 0.44 ± 0.22 | 0.13 ± 0.06 | 0.16 ± 0.28 | 0.11 ± 0.20 | 0.10 ± 0.17 | 0.21 ± 0.24 | 0.19 ± 0.20 | 0.30 ± 0.25 | 0.25 ± 0.33 | 0.41 ± 0.29 | 0.11 ± 0.08 | 0.18 ± 0.12 |
accuracy | 0.85 ± 0.08 | 0.82 ± 0.05 | 0.87 ± 0.05 | 0.83 ± 0.05 | 0.77 ± 0.07 | 0.81 ± 0.08 | 0.80 ± 0.08 | 0.83 ± 0.09 | 0.84 ± 0.09 | 0.88 ± 0.07 | 0.69 ± 0.17 | 0.74 ± 0.12 |
AUC | 0.73 ± 0.11 | 0.53 ± 0.03 | 0.53 ± 0.10 | 0.51 ± 0.09 | 0.50 ± 0.13 | 0.60 ± 0.17 | 0.57 ± 0.14 | 0.64 ± 0.17 | 0.58 ± 0.17 | 0.69 ± 0.15 | 0.53 ± 0.08 | 0.60 ± 0.13 |
# – best prediction, patch, or model corresponds to the most significant (minimal P-value) of the predicted epitopes