DEOD Uncovering the dominant effects of cancer-driver genes based on a partial covariance selection method

 

 
News
2014/12/15: Source codes and data sets are updated.
Introduction
DEOD uncovers the dominant effects of cancer-driver genes. It integrates matched copy number, gene expression and mutation data, and PPI network to construct a large scale gene network then, a scoring function is proposed to measure the effect of each gene throughout the network.
Paper Amgalan, B. and Lee, H. (2014) Uncovering the dominant effects of cancer-driver genes based on a partial covariance selection method.
Source codes
Source codes were implemented in MatLab and are available here. Instructions for running source codes are here. The source codes require seven input files such as Cancer and Normal copy number and gene expression samples in matrix forms, PPI network as a list of interactions, a list of genes for the analysis and a damaging rate matrix for mutations. Input files for breast cancer data can be downloaded from here. Note that source codes and input files should be located in same folder.
Data sets

The breast cancer data sets used in the experiments were collected from TCGA. Protein-protein interaction data sets were collected from the Human Protein Reference Database HPRD(Prasad et al.,2009) consisting of a list of interactions in the PPI network.

Contacts
bayaraa at gist.ac.kr, hyunjulee at gist.ac.kr