Tutorials

In this section, you will find examples of how to use the bulkdgd package for different tasks.

If you want an overview of the command-line utilities available in bulkdgd, see the Command-line interface section.

Here, we will showcase the usage of the package’s functions in the context of larger analysis scripts.

Specifically, we provide detailed tutorials to:

• Find the best representations in latent space for a new set of samples (Tutorial 1).

• Perform differential expression analysis between a set of “treated” samples (for instance, cancer samples) and their corresponding “untreated” samples (“normal” samples) found using the bulkdgd model (Tutorial 2).

• Train the bulkdgd model on a new set of samples (Tutorial 3).

• Download and prepare a set of samples from the Recount3 platform for use with the bulkdgd model, using a real glioblastoma dataset as an example (Tutorial 4).

• Use bulkdgd directly from R, via reticulate, to find representations and perform DEA without writing any Python code (Tutorial 5).

The data and Python notebooks needed to reproduce the tutorials can be found in the different tutorial_* directories inside the tutorials directory at the root of the repository.