Skip to main content

Functional non-coding mutations in human cancer

Final Report Summary - CANCERGENOMES (Functional non-coding mutations in human cancer)

Executive Summary
Previous studies of cancer genomes have so far been limited to sequencing and characterization of somatic DNA mutations in protein coding regions, comprising less than 2% of the human genome. Important regulatory regions such as gene promoters and mRNA non-protein-coding regions, which comprise ~94% of the transcribed DNA, are therefore still largely an uncharted black box in cancer biology. The proposed research project, “CancerGenomes”, aims to systematically detect and
functionally characterize somatic mutations in non-coding
regulatory regions using massive cancer genome datasets.

Potential impact
It is estimated that cancer causes more than 25% of all deaths in Europe (WHO 2004 estimate), and the total cost of cancer care in the EU was recently estimated to be over €100 billion annually. This represents a massive socio-economic burden on society, and there is an urgent need to improve approaches for cancer diagnosis and more effective personalized and targeted treatments. The proposed research project aims to discover and characterize new cancer driver mutations in non-coding regions, which can lead to new anti-cancer drug targets and more accurate diagnosis of cancer patients.

Project context and objectives:

Objective 1: Establish a computational infrastructure and pipeline to infer somatic mutations using paired (from same patient) cancer and normal whole genome sequences. This pipeline will use advanced computer infrastructure in order to handle the massive sequence datasets, and will incorporate existing state-of-the-art methods for individual tasks such as somatic mutation calling.

Objective 2: Develop a new statistical framework for evaluating the functional impact of non-coding mutations. Develop methods to evaluate the functional impact of non-coding mutations targeting mRNA regulatory regions, and will integrate orthogonal datasets from TCGA to directly measure if mutations confer changes in mRNA splicing and/or mRNA/protein abundance.

Objective 3: Derive a pan-cancer atlas of functional non-coding mutations. Apply the methodology in objective 1 and 2 to >700 cancer genomes from 20 cancer types in TCGA. This atlas of functional non-coding cancer mutations and targeted genes will be integrated with known cancer pathways and regulatory networks, and could lead to new diagnostic and therapeutic approaches.
Main results

The project has been terminated early after 2.5 months due to resignation of the Marie-Curie Fellow.