Skip to main content

Table 2 Comparison of predictions using different methods.

From: Computational identification of condition-specific miRNA targets based on gene expression profiles and sequence information

  SVM+SC3 SC1 SC3
TP (True Positive) rate 0.36 0.20 0.83
FP (False Positive) rate 0.00 0.00 0.03
TN (True Negative) rate 1.00 1.00 0.97
FN (False Negative) rate 0.64 0.80 0.17
Sensitivity (TP/(TP+FN)) 0.36 0.20 0.83
Specificity (TN/(TN+FP)) 1.00 1.00 0.97
Precision (TP/(TP+FP)) 1.00 1.00 0.97
  1. SVM+SC3, the method combining the SVM classifier and the scoring method based on the sequence matches. SCτ indicates the score cutoff, τ. The results were obtained with 100 test sets.