Open Collections

Galbraith, Andrew Joseph

University of British Columbia

5-methylcytosine (5mC) and 8-oxoguanine (8oxoG) are two DNA modifications which each have epigenetic roles and are implicated in disease pathophysiology. 5mC, commonly referred to as DNA methylation, is a heritable epigenetic mark that primarily silences gene expression by disrupting transcription factor binding to regulatory elements. While being a well-characterized mark, methods to detect 5mC such as whole genome bisulfite sequencing and methylation arrays are limited in scope and cannot sufficiently phase 5mC. Phasing of 5mC is necessary to measure allele-specific methylation (ASM) which is the phenomena of asymmetrical methylation profiles across different alleles. Nanopore sequencing is a long-read technology which can generate megabase scale phase blocks of 5mC. This sequencing technology was employed to identify ASM across 207 tumour genomes encompassing a variety of tumour types. Recurrent ASM was found in the context of both tumour suppressor gene (TSG) and oncogene regulation. Notably, intragenic methylation was found to be a potential regulator of cancer gene and isoform expression. In the TSGs BRCA1 and RAD51C, promoter methylation with a double hit of loss of heterozygosity (LoH) was found to result in the clinically relevant homologous recombination deficiency (HRD) phenotype. Breast and ovarian cancers with aberrant methylation were found to be enriched in somatic alterations in gene regulatory regions. A subset of HRD cases, were found to have promoter indels which may be acting as reversion mutations to restore gene functionality and confer treatment resistance. The use of nanopore resulted in an unprecedented scale and resolution to explore the regulatory mechanisms of 5mC in cancer. Reactive oxygen species (ROS) can convert guanine to 8oxoG which is a mutagenic mark that leads to G > T mutations. 8oxoG is prominent in mitochondrial disease as mitochondrial DNA (mtDNA) is particularly susceptible to the intrinsic production of ROS during aerobic respiration. Using nanopore sequencing, a recurrent convolutional neural network was trained to detect 8oxoG in the mitochondrial genome. This model was found to be over 99.9% specific and was able to distinguish samples with mitochondrial disease. Ultimately, this thesis displays the potential of nanopore to quantify and characterize DNA modifications.

Text

2024-07-05

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/88582

Bioinformatics

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/