Feature Extraction from Protein Sequence (FEPS)

The input to the webserver is a fasta-formatted protein sequence file. In a typical classification scenario, you may have protein sequences for different groups (download the tutorial). The sequences belonging to the same group are saved together in a single multiple-sequence fasta-formatted file. The input sequences have to meet following guidelines:

• The sequences must be valid protein sequences
• The sequences must be in fasta format
• The sequences of the same group are saved in one file
• The file name can represent the group name

The features are divided into 7 types. Each one contains different feature types. Select a feature type and then select corresponding feature type options from the drop-down menu.
Amino Acid Composition
Composition, transition and distribution
Autocorrelation Descriptors
Pseudo Amino Acid Composition
Quasi and sequence-order-coupling
Shannon entropy descriptors
Other descriptors

Amino acid composition Dipeptide composition Spectrum descriptors of 3-mers AA composition, Dipeptide and 3-mers

Shannon entropy Relative entropy Information gain Entropy, relative entropy, and gain

Conjoint triad

Composition Transition Distribution of AADs Composition descriptors of AADs Transition descriptors of AADs Distribution descriptors of AADs Composition descriptors based on charge of AADs Composition descriptors based on hydrophobicity of AADs Composition descriptors based on normalized VDWV of AADs Composition descriptors based on polarity of AADs Composition descriptors based on polarizability of AADs Composition descriptors based on 2ndary structure of AADs Composition descriptors based on solvent accessibility of AADs Distribution descriptors based on charge of AADs Distribution descriptors based on hydrophobicity of AADs Distribution descriptors based on normalized VDWV of AADs Distribution descriptors based on polarity of AADs Distribution descriptors based on polarizability of AADs Distribution descriptors based on 2ndary structure of AADs Distribution descriptors based on solvent accessibility of AADs Transition descriptors based on charge of AADs Transition descriptors based on hydrophobicity of AADs Transition descriptors based on normalized VDWV of AADs Transition descriptors based on polarity of AADs Transition descriptors based on polarizability of AADs Transition descriptors based on 2ndary structure of AADs Transition descriptors based on solvent accessibility of AADs

All autocorrelation descriptors (8 properties) Geary autocorrelation descriptors (8 properties) Custom Geary autocorrelation with an amio acid property Moran autocorrelation descriptors (8 properties) Moran autocorrelation descriptors with an amio acid property Normalized Moreau-Broto autocorrelation (8 properties) Custom normalized Moreau-Broto autocorrelation with an amino acid property

Type I Pseudo amino acid composition (with single or multiple AA properties) Type I first 20 Pseudo amino acid composition (with single or multiple AA properties) Type I last lamda Pseudo amino acid composition (with single or multiple AA properties) Type I Pseudo amino acid composition with physicochemical property ID Type II pseudo-amino acid compostion (hydrophobicity and hydrophilicity) Type II first 20 Pseudo amino acid composition (hydrophobicity and hydrophilicity) Type II last lamda Pseudo amino acid composition (hydrophobicity and hydrophilicity)

All quasi sequence order descriptorsl (Grantham + Schneider Wrede matrix) Quasi sequence order descriptors (select a matrix) First 20 quasi sequence order descriptors (select a matrix) Last maxlag quasi sequence order descriptors (select a matrix) All sequence order coupling numbers (Grantham + Schneider Wrede matrix) Sequence order coupling numbers (select a matrix)

Some feature types have options (see the supporting document). You may use the default options or choose options that you want. Moreover, please bear in mind that whenever 'ID Number' is an option, you can select one out of 544 Amino Acid Physicochemical properties from the drop-down menu or enter ID number to specify the amino acid physicochemical properties.

Enter your amino acid indices

A

C

E

D

G

F

I

H

K

M

L

N

Q

P

S

R

T

W

V

Y

You can choose one or more file formats. The following are the most common feature file formats accepted by machine learning packages (e.g. weka, svm-light). Whenever, the input file includes the sequences of a protein group, the last column of the output file represents the class labels.