Grant: Project Grant

Cancer is a genetic disease however most cancers are not inherited but are instead caused by random mutations in individuals' DNA over time. While most such mutations are harmless, a minority called cancer driver mutations (drivers) affect critical genes and cause unlimited cell growth and other cancerous properties. To understand cancer biology and develop therapies, we need to deeply characterize drivers. As drivers are known to affect the same DNA region more frequently than expected by chance alone, we can find drivers statistically by analysing many cancer genomes and millions of mutations. While earlier studies only mapped the small fraction (<2%) of the human genome that produces proteins, recent DNA sequencing tools map entire genomes. As such whole-genome sequencing datasets are only becoming available recently, little is known about drivers in the entire genome. We will build the largest dataset of whole cancer genomes by combining all available studies of this kind, using statistical and machine-learning tools we and others developed. We will study different cancer types and also combine all data to find drivers that are rare but important in many cancer types. We will focus on the least-studied part of the genome outside of protein-coding genes. As genes often act together in cancer, we will find additional drivers by grouping biologically related genes. We will conduct modern genetic experiments using CRISPR with cancer cell lines and confirm top-ranking candidates as true drivers of cancer by measuring cell growth, motility and other oncogenic properties. We will produce a reliable catalogue of cancer drivers supported by validation experiments of our best findings at project completion. The catalogue will increase understanding of cancer biology and ultimately improve therapies. Some drivers may be applicable as biomarkers to diagnose cancer and predict response to chemotherapies.