Skip to main content
. 2016 Jan 6;32(9):1316–1322. doi: 10.1093/bioinformatics/btw002

Table 3.

List of all target recognition features that were used for SVM target prediction modelinga

Feature name RFE rank
Seed site conservation 1
Target site location in UTR 2
GC content of target site 3
UG count 4
AG count 5
UTR length 6
Free energy of seed sequence binding 7
Pentamer motif match 8
GC content of 3′ end of target site 9
GC content of upstream nucleotides of target site 10
Position 1 A 11
Position 9 U 12
Position 1 G 13
CG count 14
Position 13 G 15
UA count 16
Position 11 G 17
AC count 18
A count 19
Position 9 C 20
Position 12 G 21
Target site <200 n.t. to UTR end 22
Position 10 AU 23
AA count 24
Position 1 AU 25
GG count 26
Position 16 G 27
GA count 28
Position 14 G 29
Position 9 A 30
Position 13 C 31
Position 9 AU 32
Position 13 AU 33
C count 34
CU count 35
AU count 36
CC count 37
Position 12 C 38
Position 11 AU 39
Position 0 G 40
Position 11 C 41
Position 15 G 42
GU count 43
Position 16 U 44
UC count 45
CA count 46
Position 16 C 47
G count 48
Target site > 800 n.t. to UTR end 49
Position 11 U 50

aTop-ranking target recognition features were identified by SVM-RFE. The seed-binding site is designated as positions 2–8. These features were summarized by analysis of target sites in 3′-UTR sequences.