Table 1 Characteristics of tools for identifying pre-miRNAs.
From: ViralmiR: a support-vector-machine-based method for predicting viral microRNA precursors
Tool | Classifier | Used features | SN (%) | SP (%) | References |
|---|---|---|---|---|---|
Triplet-SVM | SVM | Each hairpin is encoded as a set of 32 triplet elements | 93.3 | 88.1 | Xue et al. [9] |
MiPred | Random forest | 32 Triplet-SVM features and a minimum of the free energy of the secondary structure | 89.3 | 93.2 | Jiang et al. [10] |
miPred | SVM | 17 primary sequencing features, 5 secondary structural features, and 7 normalized features | 84.5 | 97.9 | Ng and Mishra [11] |
miR-KDE | RVKDE | 29 miPred features and 4 stem-loop features | 88.9 | 92.6 | Chang et al. [12] |
microPred | SVM | 29 miPred features, 4 RNAfold-related features, 6 Mfold-related features, 7 base-pair-related features, and 2 MFE-related features | 83.3 | 99.0 | Batuwita et al. [13] |
MiRenSVM | SVM | 8 triplet structural features, 8 base-pair group features, 16 thermodynamic group features | 87.7 | 98.8 | Ding et al. [14] |
miR-BAG | Naïve Bayes BF Tree SVM | 4 mononucleotide features, 16 dinucleotide features, 20 triplet structural features, consecutive paired bases, structural profile scoring, and normalized sequence-based total-pairing features | 89.8 | 91.5 | Ashwani Jha et al. [15] |