iFeatureWeb > Descriptors
iFeature is capable of calculating and extracting a comprehensive spectrum of 18 major sequence encoding schemes that encompass 53 different types of feature descriptors. iFeature also integrates 12 commonly used feature clustering, selection and dimensionality reduction algorithms in order to filter out redundant features and retain useful and relevant features.
Please note that in order to use the predicted secondary structure, protein disorder, accessible surface area, PSSM and KNN feature descriptor groups, users need to generate and provide the predicted protein property files or train files as the input file using the standalone iFeature package (See Supplementary File ). The online web server of iFeature does not support the calculation of these feature descriptors. Please download the iFeature package for calculating these descriptors.
Step 1: Input sequences in the fasta format:
Paste protein sequences in fasta format (example):
Or upload a file:
Step 2: Select Feature Descriptor(s): ("*" indicates that in order to generate the corresponding feature descriptor, the input protein/peptide sequences should have an equal sequence length)
Select output order for Group 1 descriptors:          When "userDefined" option is selected, users can specify the outout order manually.
Group 1: Amino acid composition:  Amino Acid Composition  Enhanced Amino Acid Composition *  Composition of K-Spaced Amino Acid Pairs  Tripeptide Composition
 Dipeptide Composition  Dipeptide Deviation from Expected Mean    
Group 2: Grouped amino acid composition  Grouped Amino Acid Composition  Enhanced Grouped Amino Acid Composition *  Composition of K-Spaced Amino Acid Group Pairs  Grouped Tripeptide Composition
 Grouped Dipeptide Composition      
Group 3:Autocorrelation  Moran Autocorrelation  Geary Autocorrelation  Normalized Moreau-Broto Autocorrelation  
Group 4: C/T/D  C/T/D Composition  C/T/D Transition  C/T/D Distribution  
Group 5: Conjoint Triad  Conjoint Triad  K-Spaced Conjoint Triad    
Group 6: Quasi-sequence-order  Sequence-Order-Coupling Number  Quasi-Sequence-Order    
Group 7: Pseudo-amino acid composition  Pseudo-Amino Acid Composition  Amphiphilic Pseudo-Amino Acid Composition    
Group 8: PseKRAAC ktuple value:
Lambda value:   
Gap value:   
 type-1        RAAC Cluster:  type-2        RAAC Cluster:  type-3        RAAC Cluster:  type-4        RAAC Cluster:
 type-5        RAAC Cluster:  type-6        RAAC Cluster:  type-7        RAAC Cluster:  type-8        RAAC Cluster:
 type-9        RAAC Cluster:  type-10       RAAC Cluster:  type-11       RAAC Cluster:  type-12       RAAC Cluster:
 type-13       RAAC Cluster:  type-14       RAAC Cluster:  type-15       RAAC Cluster:  type-16       RAAC Cluster:
Group 9: Binary  Binary *      
Group 10: AAindex  AAindex Descriptor *      
Group 11: BLOSUM62  BLOSUM62 Descriptor *      
Group 12: Z-scale  Z-Scale Descriptor *      
Step 3: Select Feature Clustering Algorithms (Optional):
Clustering for:  Sample Clustering       Feature Clustering Plot?:   (Notice: It will spend more time when 'Yes' is selected.)  Yes       No
Clustering algorithms:  K-Means clustering  Hierarchical clustering  Affinity Propagation clustering  Mean Shift clustering  DBSCAN clustering
Step 4: Select Feature Selection Algorithms (Optional):
Paste the label file(example):
Feature selection algorithms:  Chi-Square feature selection  Information Gain feature selection  Mutual Information feature selection  Pearson Correlation feature selection
Backend computation is powered by our Python package iFeature.