Figure 4From: Protein-RNA interface residue prediction using machine learning: an assessment of the state of the artA comparison of the IDSeq and PSSMSeq features on the RB144Seq dataset. The PSSMSeq encoding leads to a better prediction performance compared to the IDSeq encoding. (a) ROC curves and (b) PR curves showing the difference between the IDSeq and PSSMSeq features across 3 different classifiers, Naïve Bayes (NB), Support Vector Machine (SVM) with linear kernel (LK), and SVM with radial basis function (RBF) kernel.Back to article page