Examining interactions among SNPs that can explain the prognostic variability in colorectal cancer

Curtis, Aaron Albert (2023) Examining interactions among SNPs that can explain the prognostic variability in colorectal cancer. Masters thesis, Memorial University of Newfoundland.

[English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (4MB)

Abstract

Background: Colorectal cancer is a significant medical burden worldwide and in Newfoundland and Labrador. Examining the relationships of SNP interactions with survival outcomes can help identify new prognostic markers for this disease. Objectives: To examine associations between colorectal cancer survival outcomes and interactions of SNPs from MMP family and VEGF interactome genes using data-reduction methods. Methods: Two data-reduction software programs, Cox-MDR and GMDR 0.9, were applied to the data of patients from the Newfoundland Familial Colorectal Cancer Registry. Eight datasets were investigated: one for the MMP gene SNPs (201 SNPs), and seven for the VEGF interaction networks (total 1,517 SNPs). Significance of interaction models was assessed using permutation testing. Associations between significant interaction models and clinical outcomes were confirmed using multivariable regression methods. Results: For the MMP dataset two multi-SNP models and one single-SNP model were identified, while fifteen novel multi-SNP models and thirteen single-SNP models were identified for the VEGF interaction network datasets. All but one of these models were able to distinguish patients based on their outcome risk in multivariable regression models (p-value range: 0.03 – 2.2E-9). Conclusion: This research demonstrated that novel genetic interactions associated with outcome risk in colorectal cancer can be found using data-reduction methods. This proves the utility of these methods in prognostic research.

Actions (login required)

View Item