Skip to main content

Advertisement

Table 6 Performance of CLCLpred and the existing predictors tested on the ATP168 and ATP227

From: Simplified sequence-based method for ATP-binding prediction using contextual local evolutionary conservation

Dataset Method ACC TPR FPR AUC
ATP168 CLCLpred (our proposed method) 0.824 0.785 0.137 0.894
  PSSM (reference [10]) 0.752 0.700 0.196 0.823
  binary (reference [10]) 0.663 0.655 0.330 0.725
  PSSM (reference [17]) 0.757 0.757 0.243 0.841
  LogisticPSSMa (reference [17]) 0.765 0.763 0.234 0.849
  LogisticPSSM+ Bipro-aa (reference [17]) 0.770 0.769 0.228 0.855
  LogisticPSSM+ Bipro-dis (reference [17]) 0.769 0.766 0.229 0.854
  LogisticPSSM+ Bipro-sa (reference [17]) 0.772 0.770 0.225 0.856
  LogisticPSSM+ Bipro-ss (reference [17]) 0.775 0.774 0.224 0.858
  LogisticPSSM+ Bipro-allb (reference [17]) 0.772 0.770 0.227 0.857
ATP227 CLCLpred (our proposed method) 0.828 0.789 0.133 0.899
  ATPsite (reference [16]) 0.675 0.361 0.012 0.854
  Rate4site (reference [16, 26]) 0.658 0.446 0.130 0.749
  PSSM (reference [17]) 0.782 0.783 0.218 0.861
  LogisticPSSMa (reference [17]) 0.794 0.792 0.204 0.873
  LogisticPSSM+ Bipro-aa (reference [17]) 0.798 0.798 0.201 0.877
  LogisticPSSM+ Bipro-dis (reference [17]) 0.798 0.797 0.201 0.876
  LogisticPSSM+ Bipro-sa (reference [17]) 0.800 0.800 0.199 0.880
  LogisticPSSM+ Bipro-ss (reference [17]) 0.802 0.801 0.197 0.881
  LogisticPSSM+ Bipro-allb (reference [17]) 0.801 0.800 0.197 0.880