Adding a preprocessed dataset

[ShevAcharya]

Hello Pau and thanks so much for putting this pipeline together - very exciting!

I have a question regarding adding a custom dataset to the pipeline. You provide a description of how to add a new dataset from scratch, which includes steps such as downloading, peak calling, annotation, etc.

I am wondering if there's a 'shortcut' to add a custom multiome dataset that has already been processed and placed into a MuData multiome object. If you could share a snippet of code to do that, I would greatly appreciate it.

Many thanks,
Shev (Mikhail Spivakov's lab)

[PauBadiaM]

Hi @ShevAcharya ,

Sorry for the delayed response, to add a new (already processed) dataset you will need to create a annotate_<dataset_name> rule that as output produces the final dataset. In the shell you could do nothing, but this structure is still needed for the pipeline. This rule should also include the fragment files as an output.
Additionally, you still need to add the necessary metadata in the config file (config/config.yaml), pointing the dataset name, its sample names, and it's case (you can follow the other datasets as an example). Let me know if you have more questions!