Data generation

The following section outlines the data that can be generated following the training of accurate ChromBPNet models.

This is a detailed overview of the downstream functions written, available within this repo within ChromBPNet_pipeline/functions.

The types of data that can be generated include:

• Predicted accessibility tracks
• Marginal footprint predicted profiles
• Predicted DeepLIFT contribution scores
• De novo TF-MoDISco motif discovery