###############
# PRIME is a PRedictor of class I IMmunogenic Epitopes. It combines predictions of binding to HLA molecules and propensity for TCR recogntion.
#
# PRIME1.0 can be used freely by academic groups for non-commercial purposes (see license).
# The product is provided free of charge, and, therefore, on an "as is"
#  basis, without warranty of any kind.
#
# FOR-PROFIT USERS
# If you plan to use PRIME (version 1.0) or any data provided with the script in any for-profit
# application, you are required to obtain a separate  license.
# To do so, please contact [email protected] at the Ludwig Institute for  Cancer Research Ltd.
#
# If you use PRIME1.0 in a publication, please cite:
# Schmidt et al., Prediction of neo-epitope immunogenicity reveals TCR recognition determinants and provides insight into immunoediting. Cell Reports Medicine, 2021.
#
# For scientific questions, please contact David Gfeller ([email protected])
#
# Copyright (2020) David Gfeller
###############

############
INSTALLATION
############

For Mac and Linux:

1) Dowload the PRIME1.0.zip file and move it to a directory
of your choice, where you have writing permissions.

2) Unzip the PRIME1.0.zip package.

3) Open the PRIME file and replace lib_path by the
path to the PRIME1.0/lib directory where you have installed the script

4) To run it from anywhere on your computer, add the PRIME1.0 directory to your path.

5) PRIME requires MixMHCpred v2.0.2 or above to be installed on your computer (https://github.com/GfellerLab/MixMHCpred) and be in your PATH (if not in your PATH, you can use the '-mix PATH_TO_MIXMHCPRED' command line option).

6) To test your installation, make sure you are in the PRIME1.0/ directory and run the command (should not take more than a few seconds):

   ./PRIME -i test/test.txt -o test/out.txt -a A0201,A0101

The output file test/out.txt should be exatly the same as test/out_compare.txt


########
RUNNING
########

Command:
PRIME -i input_file -o output_dir -a allele1,allele2

-i input_file:
File listing all the peptides. Ideally a fasta file, but text files
are supported and lines starting with ">" are skipped.
All peptides should between 8 and 14-mers.

-a allele:
HLA-I allele. Use the simple nomenclature like A0101.
A01:01, A*01:01 or HLA-A*01:01 are also supported, but not
recommanded. If you want to make predictions with multiple
alleles, list the different alleles separated by a single comma ","
(e.g. -a A0101,A2501,B0801,B1801).

-o output_file:
The name of the output file.

Optional parameters:

-mix: Give the path of MixMHCpred executable. Version 2.1 or above is required
-predBinding: Decide whether to use MixMHCpred or NetMHCpan (default is MixMHCpred). Importantly, PRIME was trained using MixMHCpred. Although the %rank of NetMHCpan are quite similar to those of MixMHCpred and span the same range of values, PRIME predictions may be less accurate when using NetMHCpan. Using NetMHCpan also implies that NetMHCpan4.0 or above is installed in your computer and available in your path. NetMHCpan is required for predictions in non-human MHC alleles.
-thr: Threshold on %rank affinity (all peptides with %rank higher or equal to the threshold a given a score of 0 in PRIME) (default is 5).

########
Output
########

PRIME.txt provides the output of the immunogenicity predictions.
Column 1: Peptide sequence
Column 2: Lowest %Rank for PRIME score across the different alleles
Column 3: PRIME Score corresponding to alleles with the lowest %Rank
Column 4: Predicted binding score (MixMHCpred %Rank)
Column 5: Best allele
Additional columns: %Rank, scores and MixMHCpred %Rank for each allele.

As in NetMHCpan, %Rank (Column 2) indicates the percentile of randomly generated peptides that would have a score higher or equal.