Skip to main content

Table 5 Performance comparison with existing approaches for S-sulphenylation prediction on the independent test

From: SIMLIN: a bioinformatics tool for prediction of S-sulphenylation in the human proteome based on multi-stage ensemble-learning models

MethodSensitivitySpecificityMCCAccuracyAUC
SOHPRED0.730.740.34N.A.b0.80
PRESS0.680.690.2773.8%N.A.
iSulf-Cys0.730.640.3166.8%0.72
SulCysSite0.770.71N.A.72.0%0.76
SIMLIN0.880.560.3988.0%0.82
MDD–SOHa0.850.870.5887.0%N.A.
  1. aThe performance values of MDD-SOH were extracted from the study of Bui et al [6]
  2. bN.A.: not available
  3. The bold font shows the highest performance of each feature among the RF and SVM