Cancer Site/Type classification using Convolutional Neural Network
Presented by S. Ravichandran, Ph.D., BIDS, Frederick National Laboratory for Cancer Research (FNLCR)
This document will explain how to use genomic expression data for classifying different cancer/tumor sites/types. This workshop is a follow-up to the NCI-DOE Pilot1 benchmark also called TC1. You can read about the project here, https://github.com/ECP-CANDLE/Benchmarks/tree/master/Pilot1/TC1
To begin:
-
Click the launch Binder button below to begin tutorial using the dynamic versions of TC1-dataprep.ipynb and TC1-ConvNN.ipynb
-
Please note that Binder server setup on the cloud will take < 3 minutes at most. You will first see a Binder page with some log messages. After the setup, you will see an instance of Jupyer notebook in your browser. Click the Jupyter notebook, predict-drugclass.ipynb, to begin the tutorial.
-
Binder does not work with Safari on Mac OS, instead use the Chrome browser. If you are on Windows, please use Chrome.
-
If you have trouble with Binder, click either TC1-dataprep.ipynb or TC1-ConvNN.ipynb above to view a static Python JupyterNotebook.
