Deep Learning for Cancer Therapy

Motivation:

Machine learning can be used to perform unsupervised learning on the DNA sequences of patient tumors in order to identify non-linear features of the tumor DNA that help predict treatment.

Goal:

To use deep learning to identify patients whose tumor DNA mutations “look similar to” other tumors for which treatments are effective.

Platforms:

TensorFlow

Installations:

Installing Tensorflow: https://www.tensorflow.org/versions/r1.8/install/

Team & Contact:

|Suraj Jena skjena@ucdavis.edu|
|Kumud Ravisankaran kravisankaran@ucdavis.edu|
|Valeria Brewer valramirez@ucdavis.edu|
|Ninad Mehta ntmehta@ucdavis.edu|

Project no longer maintained, due to absence of cloud credits since this was a research project & proof-of-concept was complete. Contact @author if interested in a HOWTO, or getting it back up.