98 Driver Project: Data Repository

Cancer drivers are any genomic variations that contribute to the progress of a tumor. Nearly all known cancer drivers are located in the 2% coding fraction of the human genome. However the remaining 98% contains thousands of non-coding RNA (ncRNA) genes regulatory DNA elements that control gene expression, wich is a predominant function in known cancer drivers. This project aims at modelling functional constraints acting on the non-coding genome to derive a scoring system for non-coding mutations that we use for the discovery of novel cancer drivers.

The 98 Driver Project is funded by a french ITMO Cancer "Systems Biology" grant.

Supplementary files for article: A dual model for prioritizing cancer mutations in the non-coding genome based on germline and somatic events. PLoS Comp. Biol. 2015. in press.

Authors: Jia Li, Marie-Anne Poursat, Damien Drubay, Arnaud Motz, Zohra Saci, Antonin Morillon, Stefan Michiels, Daniel Gautheret

• Suppl File 1: Tables of protein-coding, small RNA and lncRNA genes overlapping hypomutated regions for each cancer type.
• Suppl File 2: R scripts
• Suppl File 3: Training sets and annotation files for RF models (large file: 150Mb)
• Suppl File 4: Genomewise SNP and SOM scores - BED format (large file: 300Mb)