Classification of Gene Mutations and Cancer-Types with BioBombe Compressed Gene Expression Features

BioBombe analysis applied to gene expression data from The Cancer Genome Atlas (TCGA) PanCanAtlas.

Then, the compressed features learned through the serial compression are fed into several machine learning algorithms.

The algorithms are used to predict mutation status (for the top 50 most mutated genes in TCGA) and cancer-type using the compressed gene expression features.

Method and results described in https://github.com/greenelab/BioBombe. We use the classification approach outlined in https://github.com/greenelab/pancancer