Open Collections

Culibrk, Luka

University of British Columbia

Genome sequencing has transformed our understanding of human genetic diseases in recent years, not least of which is Cancer. Among the genetic abnormalities commonly observed within cancer are copy number variants, alterations in the abundance of DNA, which often affect cellular function and contribute to disease. Whole-genome sequencing has allowed for high throughput examination and identification of mutations such as single nucleotide variants within cancer, while the identification of copy number variants remains comparatively difficult. The critical task for accurate identification of copy number variants from this data remains segmentation, the task of aggregating sequences of DNA abundance observations into contiguous segments of presumably constant DNA copy number. In this dissertation, we propose a novel method for performing copy number segmentation of sequenced whole cancer genomes. We apply a novel bottom-up, coarse-to-fine segmentation algorithm alongside statistical techniques to identify tumor heterogeneity and accurately perform copy number variant detection. We compare our method with a number of other methods in a variety of contexts, including fully synthetic data, resequenced cell line data, and a large cohort of sequenced metastatic cancer genomes. Next, we apply the results of our method in the analysis of chromosomal instability patterns throughout the genome. We assess genome-wide patterns of homozygous deletion and methods of measuring chromosomal instability and its numerous interfaces with tumor biology, including mutational signatures and gene dosage effects. Finally, we investigate the prospect of identifying copy number variants using long read data from oxford nanopore instruments. We present two case reports of copy number variation analysis in these data, and subsequently assess our ability to identify these variants in metastatic cancer biopsies as compared to traditional short read sequencing methods. We subsequently investigate factors influencing our ability to identify copy number events in nanopore data. In this work, we have focused on the methods and analysis of copy number variants in human cancer. The methods and analyses performed herein will assist in research concerning these mutations and their greater role in cancer and human biology.

Text

2024-04-15

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/87833

Bioinformatics

University of British Columbia

UBCV

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