Pharmacogenomics analysis of long amplicon sequences

This Snakemake workflow assigns fully phased PGx haplotypes to targeted amplicon sequencing data generated using the PacBio platforms.

Requirements

• Conda/Miniconda
• git

Installation

• Clone this repository

  • git clone https://github.com/lumc-pgx/pgx-pipe.git

• Change to the pgx-pipe directory

  • cd pgx-pipe

• Initialize submodules

  • git submodule init

• Prepare submodules

  • git submodule update

• Create a conda environment for running the pipeline

  • conda env create -n pxg-pipe -f environment.yaml

• In order to use the pipeline on a computing cluster, update your .profile to use the drmaa library:

  • echo "export DRMAA_LIBRARY_PATH=libdrmaa.so.1.0" >> ~/.profile
  • source ~/.profile

Configuration

The pipeline behavior is configured by editing config.yaml.

Execution

• Activate the conda environment created during installation
  • source activate pgx-pipe

Example:

• pipe-runner -s Snakefile -d output_dir -c config.yaml -cl cluster_settings.yaml

Modify the config.yaml and cluster_settings.yaml according to the location of your data and run parameters. The CYP2D6_locus_config.yaml is an example of locus definition file indicated in the config.yaml and should be changed according to the gene of interest.

• Note
  • The first run of the pipeline will create conda environments where required for the individual pipeline rules. This process can take some time.
  • Subsequent runs of the pipeline will re-use these environments, resulting in faster execution.

Outputs

The output of the pipeline includes the following folders

- Annotation
- Haplotyping
- Phasing
- Structural variation
- Summary
- Variant calling

Overview

Pgx-pipe combines the individual analysis modules into a meta-workflow which can be run with a single command.