scverse tutorials

On scverse.org/learn, we aim at providing a comprehensive overview of analyses that can be performed with scverse core and ecosystem packages.

To this end, this repository contains

• a registry for tutorials listed on scverse.org/learn (see tutorial-registry)
• shared tutorials that complement more specific tutorials provided by invidiual core and ecosystem packages (see docs)

Adding tutorials

If you believe a tutorial should be added to scverse.org/learn, please open an issue. We will discuss the request in the next open community meeting and potentially suggest improvements.

To be added to our website, tutorials must fulfill at least the following requirements:

• all featured packages must be scverse core or approved ecosystem packages. This does not apply to packages that are not specific to omics data analysis (e.g. pandas, seaborn).
• the notebook author agrees to maintain the tutorial in the future and is reachable via zulip.
• the notebook contains a backlink to scverse.org/learn
• the notebook is self-contained: All required example data is downloaded as part of the tutorial

You can easily check your changes to tutorials or the registry locally:

hatch run docs:build  # for tutorial notebooks
hatch run registry:validate  # for the tutorials registry

Structure of external tutorials

While we do not mandate a specific structure for tutorials, a good tutorial typically comprises the following sections:

1. General header: The tutorial should have a general header that corresponds to the analysis.
2. Brief introduction: The tutorial should introduce the package, the analysis motivation and potentially biological background.
3. Requirements to run the notebook: Special computational requirements like memory or GPUs should be specified. Any required input from other notebooks should also be listed here.
4. Package imports: All required packages should now be imported.
5. General setup: General settings such as plotting settings or ignored warnings should be set up here.
6. Data loading: Any required datasets should be loaded here. Ideally with stable links.
7. Data preprocessing: Any data preprocessing should be done here. Depending on the method this step can be skipped.
8. Package specific tutorial: The tutorial for the package should contain a healthy mix of text and code to guide the user through the analysis.
9. Link to other important tutorials/packages/sources of information: Link to any other tutorials that might be of interest or the corresponding https://sc-best-practices.org chapter.
10. References: Any referenced papers should show up in references section.
11. Acknowledgements: All contributing authors and experts should be named.