Skip to main content

Table 11 Performance on balanced plant dataset using SVM classifier on automatically collected orthologs

From: Plus ça change – evolutionary sequence divergence predicts protein subcellular localization signals

  F Div Plant 4 classes (15) F Div Plant 3 classes (61)
  AUC MCC AUC MCC
MTS 0.62±0.11 0.24±0.21 0.66±0.08 0.35±0.14
SP 0.78±0.11 0.58±0.23 N/A N/A
CTP 0.73±0.16 0.43±0.31 0.77±0.12 0.51±0.23
N-signal-free 0.80±0.14 0.72±0.20 0.81±0.09 0.67±0.13
% accuracy 60.00±9.13 66.22±10.11
  1. The 5-fold cross-validation performance of an SVM classifier using divergence features is shown for three-way classification on balanced sets of (automatically generated) plant orthologs with or without the SP class. The number of examples is given in parenthesis at top.