Epigenome-instructed pan-cancer investigation of non-coding cancer drivers

Identifying DNA mutations that drive cancer is a major goal in cancer research, critical to understanding the mechanisms of oncogenesis and to development of anti-cancer therapy. The cancer driver mutations in the protein-coding part of genome have been extensively characterized. However, the majority of the human genome is “non-coding” and relatively little is known about the involvement of non-coding mutations in cancer development.

Here, we will aim to understand how mutations in the non-coding genome can drive cancer. We have recently developed a unique framework to search for cancer driver mutations in the non-coding genome (Tomkova et al., NAR, 2022). For example, we utilise epigenome data to decipher the functional role of different non-coding regions in the given tissue. In addition, we integrate 3D long-range interaction data to understand how non-coding regions regulate cancer-relevant genes over long genomic distances. Finally, we use DNA and RNA sequencing data from cancer patients to model background mutagenesis and identify candidate non-coding cancer driver mutations.

In this project, we will expand the framework, apply it on large-scale human cancer and healthy tissue data sets, and formulate data-driven hypotheses on the general mechanisms, which will be subsequently validated experimentally. The expected output from this study is identification of a novel widespread pan-cancer mechanism of dysregulation of oncogenes and tumour-suppressor genes. This will open doors for translational applications leveraging the mechanistic knowledge gained, such as identification of novel biomarkers and therapeutic targets for precision oncology.

In this interdisciplinary project, you will learn transferable skills, including data analysis, data visualization, machine learning, statistics, high-throughput computing, bioinformatics analysis and integration of large genomic, epigenomic, transcriptomic, single-cell sequencing, and other data set.

Depending on your preferences, you can be also involved in the experimental validation and learn molecular biology techniques, such as cell culture of cancer cell lines, CRISPR-Cas9 gene editing, sequencing, etc. Alternatively, you may investigate use of more complex computational approaches for the non-coding driver discovery, including state-of-the-art deep-learning methods successfully used in other disciplines (Popel, Tomkova, et al., Nature Communications, 2020).

You will have opportunities to develop soft skills in presenting, writing, critical thinking, experimental design, networking within the Oxford scientific community and at conferences, and public engagement, in a friendly, inclusive, and supportive environment.

This project will be suprvised by Dr Marketa Tomkova and Professor Ian Tomlinson.

Please note that only applicants who have applied for a DPhil in Clinical Medicine via the University of Oxford admissions system will be shortlisted for interview.

Please quote the project ID Ludwig01 and indicate the course code RD_CM1 in your application. Please use the project details above for your research proposal.