BPReveal Documentataion

BPReveal is a suite of tools for working with sequence-to-profile models in the vein of BPNet.

It incorporates bias correction in the style of chrombpnet-lite, but with added features to regress out very complex biases.

You can get your own copy of BPReveal from the project’s GitHub page.

It also incorporates a new interpretation tool, called PISA (Pairwise Interaction Shap Analysis) that lets you view a two-dimensional map of cause and effect over a region.

For a summary of the components of BPReveal, see Programs. That section also has a list of some frequently-used API features.

Authors

  • Charles McAnany, Ph.D.

  • Melanie Weilert

  • Haining Jiang

  • Giorgio Ravanelli

  • Patrick Moeller

  • Samuel Campbell

  • Anshul Kundaje, Ph.D.

  • Julia Zeitlinger, Ph.D.

Contact

The principal contact is Charles McAnany. His Stowers address is cm2363@stowers.org, and his personal address is bpreveal@charlesmcanany.com.

At Stowers, Melanie Weilert is an excellent resource, she is the staff bioinformatician in the Zeitlinger lab.

If you’re at Stowers and want to use BPReveal for a project, the fine folks in Computational Biology are familiar with the package and can get you set up.

Indices