ORCID

https://orcid.org/0000-0001-6978-3515

Date of Award

1-11-2024

Author's School

Graduate School of Arts and Sciences

Author's Department

Biology & Biomedical Sciences (Human & Statistical Genetics)

Degree Name

Doctor of Philosophy (PhD)

Degree Type

Dissertation

Abstract

While the integration of next-generation sequencing and bioinformatics into cancer research has been immeasurably useful for advancing the field, there are still many gaps in understanding the genomic landscape of cancer due to the complexity of cancer development and evolution. Likewise, there are still many challenges for identifying and applying effective therapeutics for cancer treatment. Thus, there remains a need to continue characterizing the genomic landscape of cancer, to explore the evolution of cancer, and to understand how tumors respond to treatment. The projects described in this dissertation cover several applications of next-generation sequencing, bioinformatic analysis, and database resources for cancer research and precision oncology. The first project described involves the use of single-cell RNA sequencing, along with whole genome and whole exome sequencing, to characterize the genomics of a mouse model of bladder cancer and explore mechanisms of response to immune checkpoint blockade treatment. Differential expression and gene set enrichment analysis performed on multiple immune and stromal cell types within the tumor microenvironment revealed that IFN-g response in endothelial cells was upregulated in response to treatment, suggesting that IFN-g response in endothelial cells may play a crucial role in treatment response. Functional analysis confirmed that knocking out IFNgR1 in endothelial cells negated the treatment response observed in IFNgR1-intact mice, further indicating that IFN-g response in endothelial cells is a key mediator of effective treatment response. The second project described focuses on the use of whole genome sequencing to explore the landscape of copy number variation in over 250 pediatric brain tumors across four diagnosis groups (ATRT, Ependymoma, High-Grade Glioma, and Medulloblastoma). This analysis revealed that copy number alterations within pediatric brain tumors were quite common and could be extensive, particularly in Ependymoma, High-Grade Glioma, and Medulloblastoma. Exploration of the relationship between copy number variant (CNV) burden and overall survival suggested that CNV burden may have prognostic value within specific diagnosis groups. Likewise, analysis of the relationship between recurrently altered genomic regions and overall survival indicated that particular recurrent alterations within certain diagnosis groups could be significantly correlated with changes in overall survival. Lastly, the third project described covers updates to the Drug-Gene Interaction database (DGIdb, dgidb.org) implemented in the DGIdb 4.0 release. This resource allows researchers to explore drugs, genes, and known or predicted drug-gene interactions gathered from multiple sources in a single, harmonized database. The updates presented here (DGIdb 4.0) include the addition of several new sources, integration with crowdsource efforts, and improvements to the normalization and grouping of interactions. Collectively, this dissertation describes the use of next-generation sequencing, bioinformatic analysis approaches, and development of public database resources for cancer research that have improved the understanding and treatment of cancer.

Language

English (en)

Chair and Committee

Malachi Griffith, Obi L Griffith

Committee Members

Vivek K Arora, Megan A Cooper, Allegra A Petti, Joshua B Rubin, Ting Wang

Download

Share

COinS