The goal of immunotherapy is to demonstrate building a TensorFlow model on immunotherapy data.

This is based on the the original work by Leon Eyrich Jessen, and his blog post on RViews: “Deep Learning for Cancer Immunotherapy”.

Notebook #

This R Markdown notebook introduces the topic, analyses and plots some peptide genetic sequences.

TensorFlow model #

This deep learning model, built using Keras and TensorFlow, returns the class of a peptide. The input data is a flattened pep encoding of 180 values.

Plumber API #

To use the TensorFlow model, it is convenient to have another function that does the pre- and post-processing. This plumber API returns a peptide class given a peptide string as input.