Skip to main content

Table 9 Comparison with the best results reported for the POOL structured-based method [20] on their benchmark dataset of 160 proteins.

From: Automatic prediction of catalytic residues by modeling residue structural neighborhood

  

Performance %

 

Method/Dataset

AvgP

AvgR

AUCROC

1.

POOL(T)POOL(G)POOL(C)/allprotein (Tong et al. [20])

19.07

64.68

0.925

2.

SVM_P 24_1D1-3, 3D1-6 at Equal Precision

19.07

78.10

0.948

3.

SVM_P 24_1D1-3, 3D1-6 at Equal Recall

26.61

64.68

0.948

  1. Performance measures include: the average per-protein precision at equal recall, the average recall at equal precision, and the average area under the ROC curve (AUCROC). Results are obtained without including ligand information.