Skip to content

Latest commit

 

History

History
66 lines (48 loc) · 1.82 KB

README.md

File metadata and controls

66 lines (48 loc) · 1.82 KB

Shotgun Metagenomics Analysis Pipeline

The shotgun metagenomics pipeline deploys and runs individual software components developed at the PennCHOP Microbiome program.

Prerequisites

The deployment scripts assume that python 2.7 and pip are already installed.

Deployment

To deploy the pipeline on a fresh computer:

source deploy_virtualenv.sh
./deploy.sh

The first script will set up and activate a new virtual environment named shotgun-pipeline. Please see the notes inside the file deploy_virtualenv.sh for help with setting up virtual environment in future sessions.

The second script will then download and unzip the software and data files necessary to run the pipeline.

Testing

To test the pipeline, run the command:

./test/test_qsub.sh

This will run the pipeline on a small dataset, and should complete in a few minutes.

Usage

The pipeline expects a work directory to be set up with three files. The first file, containing the forward reads in FASTQ format, must end in _R1.fastq. The second file should contain the reverse reads in FASTQ format, and end with _R2.fastq. The third file should contain the sample names and barcode sequences, one per line, in tab-separated format. Rules for the barcode file are specified in the decontam software.

The pipeline is run by a submission script, which takes two arguments: a job ID and the work directory. For example, to run the pipeline on a dataset stored in the folder /home/norf/MyData, run the command:

/pipeline-qsub/submit.sh job1 /home/norf/MyData

Jobs submitted to the queue will then be prefixed by "job1-".

To run jobs without a queuing system, you may set the environment variable NO_QSUB to 1. See the file ./test/test_qsub.sh for an example.