Skip to main content

Table 4 Performances of the component methods and Meta-SNP on testing set

From: Collective judgment predicts disease-associated single nucleotide variants

Method Dataset Q2 PPV TPR NPV TNR MCC AUC %DB
PANTHER NSV-2012 0.74 0.81 0.71 0.68 0.78 0.49 0.75 75
PhD-SNP   0.77 0.78 0.77 0.77 0.78 0.55 0.84 100
SIFT   0.68 0.79 0.53 0.62 0.85 0.39 0.73 93
SNAP   0.64 0.59 0.91 0.80 0.38 0.34 0.79 100
CONDEL   0.75 0.78 0.70 0.72 0.81 0.51 0.82 100
Meta-SNP NSV-2012 0.79 0.79 0.80 0.80 0.79 0.59 0.86 100
  Consensus 0.87 0.88 0.89 0.87 0.85 0.74 0.91 42
  Majority 0.77 0.77 0.74 0.77 0.79 0.53 0.83 42
  Tie 0.68 0.61 0.67 073 0.69 0.35 0.72 16
  1. Q2=Overall accuracy, PPV and NPV=Positive and Negative Predicted Values, TPR and TNR=True Positive and Negative Rates. MCC=Mathew's correlation, AUC=area under the (ROC) curve. %DB is the fraction of the NSV-2012 dataset for which a prediction is returned.