Skip to content

Latest commit

 

History

History
82 lines (54 loc) · 3.84 KB

adding_your_own.md

File metadata and controls

82 lines (54 loc) · 3.84 KB
Raw

peterk87/nf-villumina: Configuration for other clusters

It is entirely possible to run this pipeline on other clusters, though you will need to set up your own config file so that the pipeline knows how to work with your cluster.

If you are the only person to be running this pipeline, you can create your config file as ~/.nextflow/config and it will be applied every time you run Nextflow. Alternatively, save the file anywhere and reference it when running the pipeline with -c path/to/config (see the Nextflow documentation for more).

A basic configuration comes with the pipeline, which loads the conf/base.config by default. This means that you only need to configure the specifics for your system and overwrite any defaults that you want to change.

Cluster Environment

By default, pipeline uses the local Nextflow executor - in other words, all jobs are run in the login session. If you're using a simple server, this may be fine. If you're using a compute cluster, this is bad as all jobs will run on the head node.

To specify your cluster environment, add the following line to your config file:

process.executor = 'YOUR_SYSTEM_TYPE'

Many different cluster types are supported by Nextflow. For more information, please see the Nextflow documentation.

Note that you may need to specify cluster options, such as a project or queue. To do so, use the clusterOptions config option:

process {
  executor = 'SLURM'
  clusterOptions = '-A myproject'
}

Software Requirements

To run the pipeline, several software packages are required. How you satisfy these requirements is essentially up to you and depends on your system. If possible, we highly recommend using either Docker or Singularity.

Please see the installation documentation for how to run using the below as a one-off. These instructions are about configuring a config file for repeated use.

Docker

Docker is a great way to run peterk87/nf-villumina, as it manages all software installations and allows the pipeline to be run in an identical software environment across a range of systems.

Nextflow has excellent integration with Docker, and beyond installing the two tools, not much else is required - nextflow will automatically fetch the nfcore/ontasfv image that we have created and is hosted at dockerhub at run time.

To add docker support to your own config file, add the following:

docker.enabled = true
process.container = "peterk87/nf-villumina"

Singularity image

Many HPC environments are not able to run Docker due to security issues. Singularity is a tool designed to run on such HPC systems which is very similar to Docker.

To specify singularity usage in your pipeline config file, add the following:

singularity.enabled = true
process.container = "shub://peterk87/nf-villumina"

If you intend to run the pipeline offline, nextflow will not be able to automatically download the singularity image for you. Instead, you'll have to do this yourself manually first, transfer the image file and then point to that.

First, pull the image file where you have an internet connection:

singularity pull --name nf-villumina.simg shub://peterk87/nf-villumina

Then transfer this file and point the config file to the image:

singularity.enabled = true
process.container = "/path/to/nf-villumina.simg"

Conda

If you're not able to use Docker or Singularity, you can instead use conda to manage the software requirements. To use conda in your own config file, add the following:

process.conda = "$baseDir/environment.yml"