DIPAN

Description

DIPAN is designed to identify neoantigens originating from intronic polyadenylation (IPA) events detected in tumor transcriptomes. These IPA-derived neoantigens have the potential to be presented by the MHC I molecules.

Installation

DIPAN incorporates Python and Bash scripts. To set up the required environment, use the provided environment.yml file to create a conda virtual environment.

1. Clone the repository:

   git clone https://github.com/YY-TMU/DIPAN.git

2. Create the environment:

   conda env create -f environment.yml
   conda activate DIPAN

3. netMHCpan and OptiType

netMHCpan can only be acquired through the official website and must be added to the environment variable after installation.

• netMHCpan

OptiType relies on Python 2.7. Due to compatibility issues with other scripts, it should be manually installed according to following instruction.

• OptiType

DIPAN options

The following options are avaliable:

• --annotated_finder/-a: Annotation file containing intron and exon information from IPAFinder.
• --bam_hla_input/-b: File containing BAM file paths and related HLA typing information.
• --bam_fq_input/-f: File containing BAM file paths and related FASTQ paths.
• --normal_proteome/-n: Amino acid sequences in the normal proteome.
• --annotated_gtf/-g: GTF file with annotated transcripts (RefSeq).
• --genome_fasta/-G: Genome in fasta format.
• --rank_threshold/-t: Set the rank threshold for high binding peptides. [default=2]
• --matched_normal/-m: Specify whether there are matched normal samples. If False, IPA-derived peptides found in normal samples are used. [default=True]
• --output_dir/-o: Output directory.
• --optitype_script: The path to 'OptiTypePipeline.py'.
• --optitype_config: Configuration file of OptiType.

Usage

1.Input

We provided the annotation file of IPAFinder for GRCh38, along with normal proteome amino acid sequences in annotated_file directory, and we recommend that users utilize it directly. The annotation file for GRCh38 of RefSeq can be downloaded from UCSC. It should be noted that the annotation file of IPAFinder must match the GTF file; if the GTF file is changed, the annotation file should be adjusted accordingly for IPAFinder.

DIPAN offers two options for users. If HLA-I typing information is unavailable, users should provide bam_fq_input, optitype_script and optitype_config. OptiType is used to calculate HLA typing. bam_fq_input should include paths to BAM files and their corresponding FASTQ files. Alternatively, if HLA typing is already known, users should provide bam_hla_input, which includes paths to BAM files along with their associated HLA typing information.

DIPAN could be tested using recommended files.

2. Unknown HLA typing

DIPAN.sh -a <IPAFinder_anno.txt> -f <bam_fq_input.txt> -n <Normal_proteome.fa> -g <refseq.gtf> -G <genome.fa> -o <output_dir> -optitype_script <OptiTypePipeline.py> -optitype_config <optitype.config>

bam_fq_input.txt contains paths of BAM file and related FASTQ file, as shown below:

Tumor  /path/tumor.sorted.bam  /path/tumor_1.fq,/path/tumor_2.fq
Normal /path/normal.sorted.bam   

3. Known HLA typing

DIPAN.sh -a <IPAFinder_anno.txt> -b <bam_hla_input.txt> -n <Normal_proteome.fa> -g <refseq.gtf> -G <genome.fa> -o <output_dir>

bam_hla_input.txt contains paths of BAM file and related HLA typing information, as shown below:

Tumor  /path/tumor.sorted.bam  HLA-A*01:01,HLA-B*44:02,HLA-C*06:02
Normal /path/normal.sorted.bam   

We collected 730 normal samples from TCGA and curated a list of IPA-derived peptides found in these normal samples. This list, along with the normal human proteome provided, can serve as a control when normal RNA-seq data are unavailable. When matched samples are missing, set matched_normal to False.

4. Output