Environment Setup¶
Gitpod¶
This material intends to be a quick hands-on tutorial on Nextflow, so we prepared a Gitpod environment with everything you need to follow it. Gitpod offers a virtual machine with everything already set up for you, accessible from your web browser or built into your code editor (eg. VSCode). To start, click on the button below.
In the gitpod window, you'll see a terminal. Type the following command to switch to the folder of this training material:
Pipeline data and scripts¶
All the files needed for the hands-on activity are stored in the directory shown below.
This includes all input data (in the data directory) and the scripts
required by the workflow (in the bin directory; see documentation).
hands-on
├── README.md
├── bin
│ └── gghist.R
├── data
│ ├── blacklist.bed
│ ├── genome.fa
│ ├── known_variants.vcf.gz
│ └── reads
│ ├── ENCSR000COQ1_1.fastq.gz
│ ├── ENCSR000COQ1_2.fastq.gz
│ ├── ENCSR000COQ2_1.fastq.gz
│ ├── ENCSR000COQ2_2.fastq.gz
│ ├── ENCSR000COR1_1.fastq.gz
│ ├── ENCSR000COR1_2.fastq.gz
│ ├── ENCSR000COR2_1.fastq.gz
│ ├── ENCSR000COR2_2.fastq.gz
│ ├── ENCSR000CPO1_1.fastq.gz
│ ├── ENCSR000CPO1_2.fastq.gz
│ ├── ENCSR000CPO2_1.fastq.gz
│ └── ENCSR000CPO2_2.fastq.gz
├── final_main.nf
└── nextflow.config
4 directories, 19 files
Script permission¶
Make sure the following R script has execute permissions:
Pulling the Docker image¶
Nextflow can pull Docker images at runtime, which is very useful as we usually work with multiple container images. The best practice when it comes to containers and Nextflow is to have a light container image for each process. This makes pulling/running/stopping containers faster and it's easier to debug, when compared to a bulky container image. Nextflow will make sure these container images are pulled, when not found locally, ran as containers, with volumes mounted plus many other things that you don't have to worry.
Even though Nextflow takes care of that for you, let’s just download manually one of the container images that we will use to see how Docker works:
It's a very lightweight container image, so the download of the layers should be fast and you will see the output below:
aca8d56a3290: Download complete
445527a00ea8: Download complete
aa5436c16a6a: Download complete
quay.io/biocontainers/samtools:1.3.1--h0cf4675_11
You can run this container and launch bash to interact with it by typing the following command:
Once inside, you can check the version of samtools, for example. You should see something like:
root@0c3770b24a24:/# samtools --version
samtools 1.3.1
Using htslib 1.3.1
Copyright (C) 2016 Genome Research Ltd.
root@0c3770b24a24:/#
Type exit to exit the container and come back to your shell.