Installing bioinformatics tools

The exercises run on your own computer or computing service. A Linux system is recommended. Windows users should use WSL2; macOS users should check each tool's platform support before starting.

Install and test the software before a timed challenge. Do not spend the challenge window resolving Docker permissions or downloading databases.

Choose one installation route

Environment	Best use
Pixi	Reproducible command-line environments, including BactScout
Conda or Micromamba	Individual Bioconda tools such as Kleborate, Snippy and IQ-TREE
Docker	Complete containerised workflows on a workstation or WSL2
Apptainer	Containers on an HPC system where Docker is unavailable
Existing institutional pipeline	Valid when its tools, versions and settings can be recorded

Avoid installing several competing versions into the same environment.

Pixi

Pixi installs a project's declared tools and dependencies.

curl -fsSL https://pixi.sh/install.sh | bash
pixi --version

For BactScout:

git clone https://github.com/ghruproject/bactscout.git
cd bactscout
pixi install
pixi run bactscout --help
pixi run bactscout version

Conda or Micromamba

Create a separate environment for each workflow. Configure Bioconda according to the Bioconda installation instructions.

conda create -n ghru-tools -c conda-forge -c bioconda \
  --strict-channel-priority snippy iqtree kleborate
conda activate ghru-tools

snippy --version
snippy --check
iqtree2 --version
kleborate --version

If your system provides Micromamba, the equivalent micromamba create and micromamba activate commands can be used.

Docker

Install Docker Engine on Linux or Docker Desktop with WSL2 integration on Windows. Confirm that containers can run:

docker version
docker run --rm hello-world

BactScout provides a ready-to-run image:

docker pull happykhan/bactscout:latest

docker run --rm \
  -v "$PWD":/data \
  --user "$(id -u):$(id -g)" \
  happykhan/bactscout:latest \
  bactscout qc /data/reads -o /data/bactscout_results

Use an explicit image tag rather than latest when you need to reproduce a completed analysis later.

Nextflow and GHRU-assembly

GHRU-assembly is a Nextflow workflow. It uses Shovill for short reads, Dragonflye for long reads and Unicycler for hybrid data.

Requirements:

Linux or WSL2;
Nextflow 24.10.3 or newer;
Docker, or an appropriate institutional container configuration;
approximately 10 CPU cores, 16 GB RAM and 50 GB free storage for the supplied examples.

Install Nextflow using the method approved by your institution, then verify it:

nextflow -version
docker version

git clone https://github.com/ghruproject/GHRU-assembly.git
cd GHRU-assembly
nextflow run main.nf --help

Hybrid and long-read execution is not currently recommended on macOS because of Medaka compatibility limitations documented by the pipeline.

Apptainer on HPC

Do not run Docker on a shared cluster unless the administrators explicitly support it. Load the institutional Nextflow and Apptainer modules, or use the versions provided by your workflow platform.

module load nextflow
module load apptainer

nextflow -version
apptainer --version

Ask the cluster team which executor, queue, project code and bind paths should be used. Test one sample before submitting the full cohort.

Record the environment

Save these details with every submission:

Operating system:
Workflow and version:
Container image and tag:
Database version:
Command:
Non-default parameters:
Analysis date:

Continue to getting started or open the guide for your exercise.