Dissertation

Searching for Microbial Nucleic Acids in Cancer Sequence Data

Gihawi, Abraham

2021

Overview

Microbes can play a prominent role in cancer: Helicobacter pylori is involved in over 90% of gastric non-cardia adenocarcinoma and Human Papillomavirus plays a major role in cervical tumorigenesis. In this thesis, I evaluate computational approaches to identify the constituent taxa within human high-throughput sequence data. I have produced the top performing approaches into a pipeline (SEPATH) which is applied to a variety of datasets: RNA sequencing of urine from patients under clinical investigation for prostate cancer and over 10,000 whole genome sequences from Genomics England's 100,000 Genomes Project. The results are sparse and rife with environmental and sequencing contaminants. Despite this, SEPATH has revealed a range of interesting bacterial and viral genera associated with tumour samples. A preliminary association is observed between the identification of specific taxa and the development of aggressive prostate cancer. Many of the genera identified have been previously suggested for association with tumours such as Bacteroides and Fusobacterium in colorectal cancer. Alphapapillomavirus was noted in oral cancer which aligns with the expected genomic features (a lack of mutations in tumour suppressor gene TP53). Also, evidence has been detected for infectious disease which will be subject to independent validation and followed up appropriately. Analysing the microbial composition of tumours could provide an additional tool to aid in therapeutic stratification of cancer patients with little added cost following sequencing.

• Publisher: ProQuest Dissertations & Theses