Skip to main content

Table 2 Selected feature set

From: Prediction of breast cancer by profiling of urinary RNA metabolites using Support Vector Machine-based feature selection

No. encoding* metabolite ratio No. encoding metabolite ratio
1 arctan (1/7)* M-1/ncm5U 23 arctan (20/3) X/ψ
2 arctan (1/12) M-1/m1A 24 arctan (20/35) X/ms2t6A
3 arctan (2/21) DHU/SAH 25 arctan (21/30) SAH/m2 2G
4 arctan (3/18) Ψ/m3U 26 arctan (22/30) m1I/m2 2G
5 arctan (5/7) C/ncm5U 27 arctan (23/18) m1G/m3U
6 arctan (5/11) C/AICA riboside 28 arctan (25/5) ac4C/C
7 arctan (6/18) acp3U/m3U 29 arctan (25/10) ac4C/m3C
8 arctan (6/19) acp3U/N6-SAR 30 arctan (25/22) ac4C/m1I
9 arctan (8/23) M-3/m1G 31 arctan (26/28) m2G/ms2io6A
10 arctan (9/21) U/SAH 32 arctan (26/30) m2G/m2 2G
11 arctan (11/5) AICA riboside/C 33 arctan (27/13) M-4/2,5-PCNR
12 arctan (11/33) AICA riboside/MTA 34 arctan (28/19) ms2io6A/N6-SAR
13 arctan (12/10) m1A/m3C 35 arctan (29/34) m2,2,7G/m6t6A
14 arctan (12/18) m1A/m3U 36 arctan (29/35) m2,2,7G/ms2t6A
15 arctan (13/9) 2,5-PCNR/U 37 arctan (30/21) m2 2G/SAH
16 arctan (13/20) 2,5-PCNR/X 38 arctan (31/2) mcm5s2U/DHU
17 arctan (14/21) m7G/SAH 39 arctan (31/24) mcm5s2U/293
18 arctan (14/26) m7G/m2G 40 arctan (33/11) MTA/AICA riboside
19 arctan (16/22) 3,4-PCNR/m1I 41 arctan (33/17) MTA/m6 1A
20 arctan (18/10) m3U/m3C 42 arctan (33/34) MTA/m6t6A
21 arctan (18/21) m3U/SAH 43 arctan (34/12) m6t6A/m1A
22 arctan (19/30) N6-SAR/m2 2G 44 arctan (34/19) m6t6A/N6-SAR
  1. Feature set for best classification performance.
  2. *Numbering of arctan encoding combinations consistent with table 1. For metabolite abbreviations, please refer to Table 1.