xiaoxiutan/pgnneo repository overview

⁠PGNneo: a ProteoGenomics-based Neoantigen Prediction Pipeline in Noncoding Regions

⁠Overview

PGNneo is a proteogenomics-based pipeline to predict neoantigens in noncoding regions. It mainly includes four modules: (1) Noncoding somatic variant calling and HLA typing; (2) Peptide extraction and customized database construction; (3) Variant peptide identification; (4) Neoantigen prediction and selection. PGNneo can be easily applied to RNA-seq data and MS data from patients of different cancer types.

⁠Installation

We provide a docker image (https://hub.docker.com/r/xiaoxiutan/pgnneo⁠) which contains all package dependencies. You need to install docker on your system in advance. Download the Dockerfile form https://github.com/tanxiaoxiu/PGNneo⁠, then the command docker build xiaoxiutan/pgnneo:v1 will pull the PGNneo image into your local machine.

docker build -t PGNneo .

⁠Usage

⁠1. Noncoding somatic variant calling and HLA typing

python model1_rnaseq_mutation_hla.py control_name_R1.fastq.gz control_name_R2.fastq.gz case_name_R1.fastq.gz case_name_R2.fastq.gz
# eg:
python model1_rnaseq_mutation_hla.py con_R1.fastq.gz con_R2.fastq.gz case_R1.fastq.gz case_R2.fastq.gz

The results of RNAseq data preprocessing, call mutation and HLA typing are in the “rna_result”, “mut_result” and “HLAtype” folders, respectively.

⁠2. Generation of tumor-specific variant peptides

python model2_mutated_peptides.py control_name_R1.fastq.gz case_name_R1.fastq.gz
# eg:
python model2_mutated_peptides.py con_R1.fastq.gz case_R1.fastq.gz

The results of the generated mutant peptides are under the “mut_result” folder.

⁠3. Database construction and variant peptide identification

python model3_MS_filtration.py

The result files will be stored under the “ms_resultmqpar/combined/txt” folder.

⁠4. Neoantigen prediction and selection

python model4_neoantigen_prediction_filtration.py

notes: Input the HLA types predicted in 3.1 or other types that the user interested in when the system prompts: "please input an HLA class I allele like 'HLA-A02:01' or multiple alleles like 'HLA-A02:01,HLA-B15:01,HLA-C01:02':".

Neoantigen predictions and filtering results will be stored in the “preneo” directory.