Skip to main content

Table 2 The 20 most important features among the 400 proposed adaptive k-skip-2-g features

From: SkipCPP-Pred: an improved and promising sequence-based predictor for predicting cell-penetrating peptides

Rank

IG(x, c)a

Features

1

0.252

RR

2

0.12

KR

3

0.119

KK

4

0.115

LR

5

0.113

MM

6

0.107

RK

7

0.107

DM

8

0.105

YM

9

0.105

ME

10

0.104

EM

11

0.103

LL

12

0.093

HM

13

0.093

DQ

14

0.092

RL

15

0.091

MH

16

0.091

DW

17

0.089

CE

18

0.088

CN

19

0.087

CM

20

0.087

GR

  1. IG(x, c)a is the information gain of feature x relative to the class attribute c. The higher the IG(x, c), the more discriminative the feature