Dissertation

Genomic Consequences of Breast and Ovarian Cancer Mutational Processes in Patient Populations and Single Cells

Funnell, Tyler

2021

Overview

Over time, cancer genomes accumulate mutations from various sources. Some mutations result from exogenous mutagens, such as UV radiation, while others reflect endogenous processes. Each source can impart characteristic patterns of mutation – mutation signatures – which can be detected using computational techniques. Loss of DNA repair mechanisms can leave specific mutation signatures in cancer cells, offering insights into tumour etiology, features for prognostic and biologic stratification and vulnerabilities to be exploited therapeutically. Whole genome sequencing methods enable detection of the full catalogue of mutations in a genome, facilitating signature discovery. To identify cancers with deficient DNA-repair, accurate methods are needed for detecting mutation signatures and their activities within individual cancers. We introduce a class of methods, topic models, that outperform standard approaches for signature analysis. We highlight a particular novel formalism for improved signature inference, based on multi-modal correlated topic models, which can at once infer signatures from both single nucleotide and structural variation counts. We show how introducing correlated structure both within and between modes of mutation can increase accuracy of signature discovery. Using 943 breast and ovarian cancers, we refine the landscape of point and structural variation signatures in these histotypes, and describe the previously poorly-understood correspondence between mutational processes in high grade serous ovarian cancer and triple negative breast cancer. We further show how mutational processes stratify these patients into clinically relevant subgroups. Our results emphasize the need to achieve a detailed picture of the historical and ongoing impacts of mutational processes in these cancers. With >26,000 single cell genomes from a subset of these cancers, we map the copy number variation properties of DNA repair deficiency with unprecedented detail. In a controlled experiment, we show that DNA repair-deficient cells have increased rates of aneuploidy and micronuclei formation. We also show that FBI and HRD cancer cells are significantly affected by different kinds of genomic instability, and describe the cell-to-cell variability of copy number breakpoints for the first time in cancer cells. Our results illustrate how mutational processes can significantly impact the manner in which genomic architectures are continually altered, and impinge upon tumour evolution.

• Publisher: ProQuest Dissertations & Theses
• Subject: Bioinformatics / Cellular biology / Oncology
• ISBN: 9798516944895