Container info

The devcontainer used in this project is based on the one used for nf-core training on nextflow, which you can find in the Nextflow training .devcontainer directory. Their latest version can be different from the currently used one in this project, so if things break, have a look there.

The devcontainer uses the nextflow-io/training container from GitHub Container Registry, as provided by the nextflow team. The Dockerfile extends this image with some additional software, e.g. for the downstream data analysis, and is built and published to the GitHub Container Registry using an action.

The devcontainer then uses this image, see, .devcontainer/codespaces-dev/devcontainer.json.

Additional software

Using an setup.sh script which is linked to the .devcontainer.json file, additional software is installed in the container. This includes:

• gdown for downloading on devcontainer startup the relevant mzML files from a private Google Drive folder (:warning: to be changed!)

• acore and vuecore to run the analysis notebook for the downstream data analysis

• set’s up conda so the notebook kernel is recognized by VSCode

Regular checks

Once a month I check that the website builds, including execution of the downstream analysis notebook 2_data_analysis.ipynb.

That the devcontainer is working is not yet automatically checked, but means manually running one.

How to update

• check for changes in nextflow-io/training repo and then set a new image tag in .devcontainer.json file

• update the setup.sh script script if needed, e.g. to install new software or new data sources