This standalone module was developed in the context of a neoantigen identification pipeline for the Immunogenomics group from Leiden University Medical Center.
-h --help Show options.
--version Show version.
-s <size(s)> List of length(s) for the sub-peptides to be generated and have binding affinity predicted.
-f <file> Three column file containing var ID, WT peptide (wild type) and MUT peptide (mutant).
-d <path/to/db> Path to database.
-b <HLA(s)> HLA alleles for the binding prediction step.
-o <output_dir> Output folder where files and sub-folders will be created.
-n <sample_name> Sample name. Will be used in the file names.
binding_prediction.py -f long_peps.txt -s 8,9 -d /path/to/db -b A*02:01,hla-a0101 -o /path/out/dir -n Sample_X
1- (Optional) Create & activate conda environment with python 3.6 and docopt:
conda create -n <ENV_NAME_HERE> python=3.6 docopt
conda activate <ENV_NAME_HERE> OR source activate <ENV_NAME_HERE>
2- Setup modified mhctools:
pip install git+https://github.com/Amfgcp/mhctools
NOTE: mhctools requires most of the binding predictors to be installed locally (see: https://pypi.org/project/mhctools/).
3- Clone this repository:
git clone https://github.com/Amfgcp/NeoSeq_WDL.git
4- Create output folder, e.g.:
mkdir test-ouput
5- Run example
python neoseq_wdl/scripts/binding-prediction/binding_prediction.py -n 1207.test -f neoseq_wdl/scripts/binding-prediction/1207.test.txt -s 8 -d /exports/path-demiranda/usr/amfgcp/databases/ncbi/v5/generated/swissprot_taxid_9606/swissprot_taxid_9606 -b HLA-C02:02 -o test-output
6- Expected output folder structure
test-output/
./25aa
1207.test_25aa_swissprot_taxid_9606.fsa
./51aa
1207.test_2massSpec_swissprot_taxid_9606.fsa
./bind-pred
1207.test_binding_prediction_swissprot_taxid_9606.txt
./blast
1207.test_blast_output_8mers_swissprot_taxid_9606.xml
1207.test_peptides_to_blast_8mers_swissprot_taxid_9606.fsa
./logs
bind-pred_1207.test_12-08-2019.log