Table 5 The top 22 important features selected by attribute weighting feature selection method for general dataset.

From: PrESOgenesis: A two-layer multi-label predictor for identifying fertility-related proteins using support vector machine and pseudo amino acid composition approach

order

Descriptor

Protein feature

Feature group

1

S

Serine

Amino Acid Composition

2

I

Isoleucine

Amino Acid Composition

3

IA

Dipeptide Composition (Isoleucine-Alanine)

Amino Acid Composition

4

solventaccess.Group1

Solvent Accessibility attribute of Composition

CTD

5

solventaccess.Group3

Solvent Accessibility attribute of Composition

CTD

6

Schneider.Xr.S

QSO in QSOD using Schneider-Wrede distance

Quasi-sequence-order

7

Grantham.Xr.I

QSO in QSOD using normalized Grantham chemical distance

Quasi-sequence-order

8

Grantham.Xd.1

QSO in QSOD using normalized Grantham chemical distance

Quasi-sequence-order

9

prop7.Tr2332

Solvent Accessibility attribute of Transition

CTD

10

prop5.G2.residue0

Charge attribute of Distribution

CTD

11

prop5.G2.residue25

Charge attribute of Distribution

CTD

12

prop5.G2.residue50

Charge attribute of Distribution

CTD

13

prop5.G2.residue75

Charge attribute of Distribution

CTD

14

prop5.G2.residue100

Charge attribute of Distribution

CTD

15

VS333

Conjoint Triad

Conjoint Triad

16

prop2.G1.residue0

Normalized van der Waals Volume attribute of Distribution

CTD

17

prop2.G1.residue25

Normalized van der Waals Volume attribute of Distribution

CTD

18

prop2.G1.residue50

Normalized van der Waals Volume attribute of Distribution

CTD

19

prop2.G1.residue75

Normalized van der Waals Volume attribute of Distribution

CTD

20

prop2.G1.residue100

Normalized van der Waals Volume attribute of Distribution

CTD

21

Schneider.Xr.I

QSO in QSOD using Schneider-Wrede distance

Quasi-sequence-order

22

Grantham.Xr.S

QSO in QSOD using normalized Grantham chemical distance

Quasi-sequence-order

Back to article page