🎉 ImmunoTyper-SR 🧬

ImmunoTyper-SR is a powerful tool for Immunoglobulin Variable Gene genotyping and CNV analysis from whole genome sequencing (WGS) short reads using ILP Optimization. Check out our paper here for more details.

📢 New Feature: Now supporting IGLV and TRAV genotyping! 🎉

🚀 Installation

Gurobi

ImmunoTyper-SR leverages the Gurobi solver for optimization. You need a valid license to use Gurobi. Licenses are free for academic purposes.

Docker / Singularity

For the easiest installation, we recommend using the Docker image available on DockerHub at cdslsahinalp/immunotyper-sr.

To run the image with Singularity (commonly used on HPCs), use the following command:

singularity pull docker://cdslsahinalp/immunotyper-sr
singularity run -B <GUROBI_LICENSE_PATH>:/opt/gurobi/gurobi.lic -B <BAM_DIRECTORY>:<BAM_DIRECTORY> -B <OUTPUT_PATH>:/output immunotyper-sr_latest.sif <BAM_DIRECTORY>/<BAM_FILE>

Conda + Pip

If you already have BWA installed and prefer not to create a new environment, you can download the latest release binary (see right toolbar) and install it with pip:

pip install <binary.whl>

For the best experience, we recommend setting up a clean environment first:

conda create -n immunotyper-SR -c bioconda python=3.8 bwa samtools
conda activate immunotyper-SR
pip install <binary.whl>

Environment and Dependencies

Installing ImmunoTyper-SR with pip will automatically install these dependencies:

• biopython
• dill
• gurobipy
• logbook
• pysam

In addition to the above, you will need

1. BWA mem mapper. We recommend using a new conda environment for the installation, which you can also use to install BWA:

conda create -n immunotyper-SR -c bioconda python=3.8 bwa samtools
conda activate immunotyper-SR
pip install <binary.whl>

2. Gurobi solver configured with a valid license

To check that gurobi is correctly configured, run gurobi_cl from a shell.

Installing from source

If the binary fails to install, you can build the tool from source:

conda create -n immunotyper-SR -c bioconda python=3.8 bwa samtools
conda activate immunotyper-SR
git clone git@github.com:algo-cancer/ImmunoTyper-SR.git ./ImmunoTyper-SR
cd ImmunoTyper-SR
python -m pip install --upgrade  build
python -m build
pip install dist/<.tar.gz or .whl build>

🛠️ Running ImmunoTyper-SR:

After installing with pip, use the command immunotyper-SR. The only required input is a BAM file. Outputs are generated in the current working directory, where is the input BAM filename without the extension:

• -<gene_type>_functional_allele_calls.txt: List of functional alleles called.
• -<gene_type>_allele_calls.txt: Includes pseudogenes.
• -<gene_type>-extracted.fa: Reads extracted from the BAM used for analysis.
• -<gene_type>-immunotyper-debug.log: Log file.

IMPORTANT: If your BAM was mapped to GRCh37 use the --hg37 flag.

$ immunotyper-SR --help
usage: immunotyper-SR [-h] [--gene_type {ighv,iglv,trav,igkv}] [--output_dir OUTPUT_DIR] [--ref REF] [--hg37] [--bwa BWA] [--max_copy MAX_COPY] [--landmarks_per_group LANDMARKS_PER_GROUP] [--landmark_groups LANDMARK_GROUPS] [--stdev_coeff STDEV_COEFF] [--seq_error_rate SEQ_ERROR_RATE] [--solver_time_limit SOLVER_TIME_LIMIT] [--debug_log_path DEBUG_LOG_PATH]
                      [--write_cache_path WRITE_CACHE_PATH] [--threads THREADS] [--no_coverage_estimation]
                      bam_path

ImmunoTyper-SR: Ig Genotyping using Short Read WGS

positional arguments:
  bam_path              Input BAM file

optional arguments:
  -h, --help            show this help message and exit
  --gene_type {ighv,iglv,trav,igkv}
                        Specify which genes to target
  --output_dir OUTPUT_DIR
                        Path to output directory. Outputs txt file of allele calls with prefix matching input BAM file name.
  --ref REF             Path to the reference FASTA to decode CRAM files. Option is not used if bam_path is not a CRAM.
  --hg37                Flag if BAM mapped to GRCh37 not GRCh38
  --bwa BWA             path to bwa executible if not in $PATH
  --max_copy MAX_COPY   Maximum number of allele copies to call
  --landmarks_per_group LANDMARKS_PER_GROUP
                        Number of landmarks per group to use (default = 6)
  --landmark_groups LANDMARK_GROUPS
                        Number of landmark groups to use (default = 6)
  --stdev_coeff STDEV_COEFF
                        Standard deviation scaling coefficient (default = 1.5)
  --seq_error_rate SEQ_ERROR_RATE
                        Expected sequence error rate (default = 0.02)
  --solver_time_limit SOLVER_TIME_LIMIT
                        Time limit for ILP solver in hours
  --debug_log_path DEBUG_LOG_PATH
                        Path to write log
  --write_cache_path WRITE_CACHE_PATH
                        Specific location and name of allele db sam mapping cache
  --threads THREADS     Max number of threads to use
  --no_coverage_estimation
                        Disables empirical coverage