pombredanne / immuno Goto Github PK

Immunotherapy Pipeline

Status

Usage

From peptide string

python mutation_report.py --string EDLTVKIGDFGLATEKSRWSGSHQFEQLS --hla "HLA-B*35:01"

From input file

python mutation_report.py --input-file <.vcf, .maf, .tab, .fasta> --hla-file <allele-file> 

If VCF file is input, it will be annotated w/ gene information and Ensembl transcripts, otherwise input an annotated SnpEFF VCF file.

If you don't either an hla or hla-file then the default "HLA:A*02:01" will be used.

Options:

• --input-file: Input file containing somatic variants (one of MAF, VCF, or SnpEff annotated VCF)
• --string: String of amino acids, with optional annotation of a mutated region between underscores (i.e. QYS_LL_Q)
• --hla: Comma separated list of HLA alleles.
• --hla-file: Text file containing one HLA allele per line.
• --peptide-length: Length of vaccine peptide (window around mutation, default 31)
• --min-peptide-padding: Minimum number of wildtype residues before or after a mutation
• --print-epitopes: Print dataframe with epitope scores
• --print-peptides: Print dataframe with vaccine peptide scores
• --output-epitopes-file: CSV output file for dataframe containing scored epitopes
• --output-report-file: Path to HTML report containing scored vaccine peptides and epitopes
• --random-mhc: Random values instead for MHC binding prediction
• --iedb-mhc: Use IEDB's web API for MHC binding
• --skip-mhc: Don't predict MHC binding
• --quiet: Suppress verbose output

Requirements

• datacache
• epitopes
• nose
• NumPy
• pandas
• BioPython

You must either have netMHCpan installed locally or pass the iedb-mhc flag to use a web-based MHC binding predictor.