Software for designing epitope based vaccine
Software Name Description Usage
PROPRED1 The ProPred-I software is useful for identification of MHC Class-I binding regions in antigens. This is helpful for the users in identifying the promiscuous regions. It implements matrices for 47 MHC Class-I alleles, proteasomal and immunoproteasomal models. propred1.pl -i input_file -o output -p path
-i Sequence in FASTA format
-o name of out put folder
nHLApred This is a comprehensive method for prediction of MHC binding peptides or CTL epitopes of 67 MHC alleles. The prediction for 30 alleles is based on the hybrid approach of Artificial Neural Networks (ANNs) and Quantitative Matrices (QM). The prediction for rest 37 MHC alleles is based on the quantitative matrices. The predicted MHC binders are filtered to potential CTL epitopes by using Proteasomal matrices. nhla4pred.pl -i input_file -o output -p path
-i Sequence in FASTA format
-o name of out put file
Pcleavage A SVM based method for prediction of protesosomal cleavage site in a protein sequence. pcleavage.pl -i input_file -o output -p path
-i Sequence in FASTA format
-o name of out put file
TAPPred TAPPred is an on-line service for predicting binding affinity of peptides toward the TAP transporter.The prediction of TAP binding peptides is crucial in identifying the MHC class-1 restricted T cell epitopes. The Prediction is based on cascade SVM, using sequence and properties of the the amino acids. The correlation coefficient of 0.88 was obtained by using jack-knife validation test. tabpred.pl -i input_file -o output -p path
-i Sequence in FASTA format
-o name of out put file
CTLPred CTLPred is a direct method for prediction of CTL epitopes crucial in subunit vaccine design.In direct methods the information or patterns of T cell epitopes instead of MHC binders were used for the development o f methods. The methods is based on elegant machine learning techniques like a Artificial Neural network and support vector machine . The methods also allows the consensus and combined prediction based on these two approaches. ctlpred.pl -i input_file -o output -p path
-i Sequence in FASTA format
-o name of out put file
ProPred The aim of this server is to predict MHC Class-II binding regions in an antigen sequence, using quantitative matrices. The server will assist in locating promiscuous binding regions that are useful in selecting vaccine candidates. propred.pl -i input_file -o output -p path
-i Sequence in FASTA format
-o name of out put file
HLA-DR4 Pred The HLA-DR4Pred is an SVM and ANN based HLA-DRB1*0401(MHC class II alleles) binding peptides prediction method. The accuracy of the SVM and ANN based methods is ~86% and ~78% respectively.The performence of the methods was tested through 5 set cross-validation. The training of SVM and ANN was done by using the freely availaible SVM_LIGHT and SNNS packages respectively. The data for training of neural network and SVM model has been extracted from MHCBN database.This method will be useful in cellular immunology, Vaccine design, immunodiagnostics, immunotherapeuatics and molecular understanding of autoimmune susceptibility. hladr4pred.pl -i input_file -o output -p path
-i Sequence in FASTA format
-o name of out put file
MHC The X-ray crystal structure of MHC class II molecule has revealed open peptide binding groove. A peptide bound in this groove may flank from one or the other side. Understanding which residues are acctually involved in binding will be very useful for understanding MHC peptide interactios. We have used Matrix Optimization Technique to predict MHC binding core. Using binders from MHCPEP and nonbinder Data with MOT an accuracy of correct classification from 97 to 99% was obtained with HLA-DR1, HLA-DR2 and HLA-DR5 allele. This is the highest accuracy reported by any method. The prediction method used in this server is based on MOT and relies on the thought that binders have unique patterns which can be easily distinguished from nonbinders. mhc.pl -i input_file -a allele -o output
-i Sequence in FASTA format
-a: Select Allele: 1 for HLA-DR1, 2 for HLA-DR2, 3 for HLA-DR5
-o name of out put file
MHCBench The MHCBench is an interface developed for evaluating the Major Histocompatibility Complex (MHC) binding peptide prediction algorithms. It allows the users to compare the performance of the old/new prediction methods in terms of the threshold dependent and independent parameters. The MHCBench offers a collection of defined data sets. The users can download these sets and can evaluate their own algorithms. The evaluations can be compared with other prediction methods evaluated using the same data sets. The server can be extended to include new methods for different MHC allele and new data sets.
Overall, the MHCBench is an effort to benchmark the prediction methods in terms of the data sets and the evaluation parameters. mhcbench.pl -i input_file -o output
-i Sequence in CSV format
-o name of out put file
ABCPred The aim of ABCpred server is to predict B cell epitope(s) in an antigen sequence, using artificial neural network. This is the first server developed based on recurrent neural network (machine based technique) using fixed length patterns. abcpred.pl -i input_file -t threshold -w window size -o output
-i: input file in fasta format
-t: Threshold Value (ranges from 0.1 to 1, Default is 0.5)
-w: Window size (Default is 16)
-o: output file
CBTOPE It has been observed that conformational B cell epitopes (~90% of all B cell epitopes) are more complex and hard to define than sequential epitopes. Several methods do exist for the prediction of conformational B cell epitope but they require antigen 3D structure or homology based model of the amino acid sequence. So far no method is available which can predict conformational B cell epitope using antigen primary sequence in the absence of any homology with the known structures. In the present study using amino acid composition as an input feature for Support vector machine (SVM) we developed a model with prediction accuracy of more than 85% and Area under curve (AUC) 0.9. cbtope.pl -i input_file -o output
-i: input file in fasta format
-o: output file LBTOPE Predict of B-cell epitopes (antigenic region) with high accuracy is one of the major challenges in designing subunit/peptide vaccine or immunotherapy. In order to overcome limiations of existing methods, we developed a method LBtope for predicting linear B-cell epitopes. We developed several models using various techniques (e.g., SVM, IBk)on a large dataset of B-cell epitopes and non-epitopes (12063 epitopes and 20589 non epitopes obtained from IEDB database). First time, experimetally valiadted non B-cell epitopes were used for developing prediction model. cbtope.pl -i input_file -o output
-i: input file in fasta format
-o: output file
