MHC class I peptide binding prediction tools - version 3.1.6
===========================================================

Release Note
------------
2024-07-10
Update allele info to support dog alleles for method NetMHCpan 4.1 (BA and EL).
2023-09-19
Two IEDB recommended methods now exist, one for the prediction of peptide binding and one for the prediction of epitopes.
2022-12-21
Fix issue regarding netmhccons missing percentile data of 22 alleles
2022-11-22
update binaries for patching vulnerabilities
2021-02-01
fix an issue regarding Duplicate Entries when multiple alleles selected
2020-11-12
fix a documentation issue regarding available MHC for mouse
2020-09-30
* update for adding a new version of prediction method NetMHCpan 4.1 (BA and EL).
2020-05-12
* This release will run only under Python 3.6 or higher.
* The 'IEDB recommended' method was updated to use NetMHCpan 4.0 EL across all alleles as of April 2020.  For more information, see the help page: http://tools.iedb.org/mhci/help/#Method .
* Loading time has been optimized, so predictions should run faster than earlier versions.
* Path validation added to configure script to ensure 'netmhccons' will work properly.



Introduction
------------
This package contains a collection of prediction tools for peptides binding to Major Histocompatibility 
Complex (MHC) class I molecules. The collection is a mixture of pythons scripts and linux 64-bit environment 
specific binaries.


Prerequisites:
-------------

+ Linux 64-bit environment
  * http://www.ubuntu.com/
    - This distribution has been tested on Linux/Ubuntu 64 bit system.

+ Python 3.6 or higher
  * http://www.python.org/

+ tcsh
  * http://www.tcsh.org/Welcome
    - Under ubuntu: sudo apt-get install tcsh

+ gawk
  * http://www.gnu.org/software/gawk/
    - Under ubuntu: sudo apt-get install gawk


Installation:
------------
Unpack the tar.gz files (IEDB_MHC_I-VERSION.tar.gz)
Run the 'configure' script to set up path variables for trained models.
            
Specifically for mhc_i:
  $ tar -zxvf IEDB_MHC_I-VERSION.tar.gz
  $ cd mhc_i
  $ ./configure

 * Please make sure the path inluding "/mhc_i" is less than 57 characters long to make method NetMHCcons working.

  
Available prediction methods are:
---------------------------------
ann
comblib_sidney2008
consensus
netmhccons
netmhcpan_ba (IEDB_recommended_binding)
netmhcpan_el (IEDB_recommended_epitope)
netmhcstabpan
pickpocket
smm
smmpmbec


Method versions used in the tool:
---------------------------------
ann                4.0
comblib_sidney2008 1.0
consensus          2.18
netmhcpan_ba       4.1
netmhcpan_el       4.1
netmhccons         1.1
netmhcstabpan      1.0
pickpocket         1.1
smm                1.0
smmpmbec           1.0


Available commands: 
------------------
* List all available commands.
./src/predict_binding.py

* List all available MHC-I prediction methods.
./src/predict_binding.py method

* List all available (MHC,peptide_length) for a given method.
./src/predict_binding.py [method] mhc

* Make predictions given a file containing a list of sequences.
./src/predict_binding.py [method] [mhc] [peptide_length] [input_file]

* Make predictions given multiple alleles/lengths. (Please note that Number of alleles and corresponding lengths must be equal.)
./src/predict_binding.py [method] [mhc1,mhc2,mhc3...] [peptide_length1,peptide_length2,peptide_length3...] [input_file]

* Make predictions given a file containing a list of sequences and user-provided MHC sequence.
** Only netmhcpan has this option.
./src/predict_binding.py [method] -m [input_file_mhc] [peptide_length] [input_file]


Example:
--------
* List all alleles for ANN method
./src/predict_binding.py ann mhc

* To use ANN method to predict binding for allele HLA-A*02:01 and length 9 with protein sequence(s) in file input_sequence.fasta
./src/predict_binding.py ann HLA-A*02:01 9 ./examples/input_sequence.fasta

* To use SMM method to predict binding for allele HLA-A*01:01 with length 9 and allele HLA-A*02:01 with length 10
./src/predict_binding.py smm HLA-A*01:01,HLA-A*02:01 9,10 ./examples/input_sequence.fasta

* To use NetMHCpan method to predict binding for user defined MHC protein sequence and length 9 with protein sequence(s) in file input_sequence.fasta
./src/predict_binding.py netmhcpan_ba -m ./examples/protein_mhc_B0702.fasta 9 ./examples/input_sequence.fasta
./src/predict_binding.py netmhcpan_el -m ./examples/protein_mhc_B0702.fasta 9 ./examples/input_sequence.fasta
./src/predict_binding.py IEDB_recommended_epitope -m ./examples/protein_mhc_B0702.fasta 9 ./examples/input_sequence.fasta

* You may also redirect (pipe) the input file into the script.
echo -e ./examples/input_sequence.fasta | ./src/predict_binding.py ann HLA-A*02:01 9
echo -e ./examples/input_sequence.fasta | ./src/predict_binding.py netmhcpan_ba -m ./examples/protein_mhc_B0702.fasta 9

* To redirect the output to a file, use:
./src/predict_binding.py smm "HLA-A*02:01" 9 ./examples/input_sequence.fasta > output.txt