The goal of this project is to provide an easy way to package, distribute and run tools and pipelines using docker images.

There are three main components:

• Format and protocol specification.
• Tools to run apps and pipelines.
• An SDK to easily package tools and pipelines into reusable components.

Note that this project is still under development. You can run pipelines, but can't (easily) make them yet.

First, install docker on a linux machine.

Second, install rabix via pip:

$ pip install git+https://github.com/rabix/rabix

Check if everything works by installing an example pipeline:

$ rabix install https://s3.amazonaws.com/boysha/pipeline_test_bwa_freebayes.json

The "install" command simply pre-fetches referenced docker images. If you don't want to download hundreds of megabytes, you can just attempt to run the pipeline, which should fetch only the JSON files and present you with input options:

$ rabix run https://s3.amazonaws.com/boysha/pipeline_test_bwa_freebayes.json 

Optionally, run the pipeline with some example data. It won't take long:

$ rabix run https://s3.amazonaws.com/boysha/pipeline_test_bwa_freebayes.json \
  --reference https://s3.amazonaws.com/boysha/testfiles/example_human_reference.fasta \
  --read https://s3.amazonaws.com/boysha/testfiles/example_human_Illumina.pe_1.fastq \
  --read https://s3.amazonaws.com/boysha/testfiles/example_human_Illumina.pe_2.fastq