Table 3 Top 21 features in human predictive model. Hairpin structural features are labeled in italics, while read compositional features are not
From: miRge 2.0 for comprehensive analysis of microRNA sequencing data
Rank | Feature name | Description of the feature |
|---|---|---|
1 | count_bindings_in_miRNA | Number of bindings in the stable range of sequences |
2 | exactMatchRatio | The proportion of reads that are an exact match to the cluster sequence in the cluster |
3 | pair_state_No | Whether there is another stable range of sequences located at the other arm of precursor |
4 | mFE | Minimum free energy (MFE) of the precursor |
5 | head_minus3_TemplateNucleotide_percentage | Proportion of genomic templated nucleotide at position −3 relative to the 5′ end of the stable range of the cluster sequences |
6 | hairpin_count | Number of hairpin loops in the precursor |
7 | stem_length | Stem length of the precursor |
8 | distanceToloop | Distance between the stable range of sequences and the terminal loop |
9 | percentage_PairedInMiRNA | Number of bindings in the stable range of sequences divided by its length |
10 | headUnstableLength | 5′ unstable length of the cluster |
11 | pair_state_Yes | Whether there is another stable range of sequences located at the other arm of precursor |
12 | tail_plus2_A_percentage | Proportion of non-templated adenine (A) at position + 2 relative to the 3′ end of the stable range of the cluster sequences |
13 | head_minus2_TemplateNucleotide_percentage | Proportion of genomic templated nucleotide at position −2 relative to the 5′ end of the stable range of the cluster sequences |
14 | binding_count | Number of bindings in the precursor hairpin |
15 | tail_plus1_A_percentage | Proportion of non-templated adenine (A) at position + 1 relative to the 3′ end of the stable range of the cluster sequences |
16 | armType_loop | Whether the stable range of sequences is located at the terminal loop if the precursor |
17 | tail_plus3_A_percentage | Proportion of non-templated adenine (A) at position + 3 relative to the 3′ end of the stable range of the cluster sequences |
18 | tail_plus5_TemplateNucleotide_percentage | Proportion of genomic templated nucleotide at position + 5 relative to the 3′ end of the stable range of the cluster sequences |
19 | tail_plus1_TemplateNucleotide_percentage | Proportion of genomic templated nucleotide at position + 1 relative to the 3′ end of the stable range of the cluster sequences |
20 | interiorLoopCount | Number of interior loops in the precursor |
21 | head_minus1_TemplateNucleotide_percentage | Proportion of genomic templated nucleotide at position −1 relative to the 5′ end of the stable range of the cluster sequences |