Table 1. Summary of features used for classification by RFMirTarget.

	Feature Name	Top 12		Feature Name	Top 12
1	Alignment score	*	18	Position 10
2	Alignment length		19	Position 11
3	Minimum free energy of the alignment	*	20	Position 12
4	G:C’s absolute frequency in the alignment	*	21	Position 13
5	A:U’s absolute frequency in the alignment	*	22	Position 14
6	G:U’s absolute frequency in the alignment		23	Position 15
7	Number of gaps in the alignment		24	Position 16
8	Number of mismatches in the alignment		25	Position 17
9	Position 1		26	Position 18
10	Position 2	*	27	Position 19
11	Position 3		28	Position 20
12	Position 4	*	29	Minimum free energy of the seed	*
13	Position 5		30	G:C’s absolute frequency in the seed	*
14	Position 6	*	31	A:U’s absolute frequency in the seed	*
15	Position 7	*	32	G:U’s absolute frequency in the seed	*
16	Position 8		33	Number of gaps in the seed
17	Position 9		34	Number of mismatches in the seed

^*The top 12 features refer to those features with greatest impact in the predictive accuracy of the RF model, estimated by means of a restricted forward feature selection.