Skip to main content

Table 11 Ablation analysis of base learner interactions in cascade meta classifiers

From: A meta-learning approach for B-cell conformational epitope prediction

Classifier* TPR FPR Precision Accuracy F-score MCC AUC
Cascade 0.199 0.008 0.519 0.958 0.288 0.304 0.789
\AAP 0.164 0.010 0.423 0.955 0.237 0.244 0.774
\ElliPro 0.144 0.009 0.414 0.955 0.214 0.226 0.748
\Bpredictor 0.149 0.009 0.435 0.956 0.222 0.237 0.743
\SEPPA 2.0 0.065 0.003 0.500 0.957 0.115 0.169 0.698
\BCPREDS 0.025 0.005 0.179 0.954 0.044 0.052 0.695
\DiscoTope 2.0 0.045 0.004 0.321 0.955 0.079 0.107 0.684
\BepiPred 0.065 0.006 0.317 0.954 0.107 0.127 0.653
\ABCpred 0.060 0.007 0.279 0.953 0.098 0.112 0.648
  1. *Classifiers tested in the ablation analysis. The first classifier in the first row is the cascade meta classifier that employs all of the 8 base learners (Figure 4). The remaining classifiers are listed in the order in which they were selected to be removed iteratively from the cascade meta classifier for the ablation study. `\’ indicates removed. For example, the second classifier is the cascade meta classifier after AAP was removed, and the third classifier is the cascade meta classifier after AAP and ElliPro were removed from the meta model. The meta classifier in the final row did not apply any base learner after the final prediction tool ABCpred was removed.