Characterization and Biomarker Discovery of Cancer-associated Novel Isoform via Hybrid Sequencing

透過混合測序技術進行癌症相關新型異構體的表徵和生物標誌物發現

Student thesis: Doctoral Thesis

View graph of relations

Author(s)

Related Research Unit(s)

Detail(s)

Awarding Institution
Supervisors/Advisors
Award date15 Oct 2024

Abstract

While it has been recently reported that cancer-specific isoforms can be used as novel prognostic biomarkers and serve as an untapped source of drug targets for immuno-oncology, the vast majority of cancer-specific isoforms remain unknown. Until now, at least 95% of genes with multiple isoforms remain to be discovered. The work in this thesis focuses on the proteogenomic characterization of cancer-associated novel isoforms (NIs) cataloged by the hybrid sequencing paradigm and the discovery of biomarkers based on these molecules. The contents of each chapter are summarized here:

Chapter 1: In this chapter, we introduced the necessity of cancer transcriptomic analysis in isoform granularity and the hybrid sequencing method to achieve isoform-level study. We first brought forward the advantages of the transcriptomic analysis in isoform granularity and summarized the recent progress of isoform-level translational research in multiple cancer types. Afterward, we compared the technical pros and cons of short- and long-read sequencing methods. The hybrid sequencing paradigm was introduced hereupon.

Chapter 2: In this chapter, we performed integrative proteogenomic characterization of cancer-associated novel isoforms in hepatocellular carcinoma. These novel isoforms served as a reservoir for downstream biomarker discovery. Three tissue transcriptome cohorts were used to identify an NI signature distinguishing HCC from normal tissue. The robustness of this signature was supported by a pan-cancer liquid biopsy cohort. Additionally, multiple NIs were found to be associated with the prognosis of HCC patients. Collectively, NIs can be an alternative source of cancer biomarkers.

Chapter 3: In this chapter, a similar characterization was conducted for breast cancer based on tissue samples. A streamlined breast cancer transcript was constructed and verified by multi-omics analysis. Transcriptomic features were examined in aging breast cancer patients, revealing insights beyond gene-level studies. New transcripts' clinical value was explored, and predictive models for chemosensitivity were developed and validated in patient-derived organoids.

Chapter 4: This chapter utilized the hybrid sequencing paradigm to characterize the transcriptome in PDAC. Our analysis built a reference PDAC transcriptome and revealed a novel isoform repertoire with widespread functional diversity, thousands of tumor-specific aberrant alternative splicing events, and prognostic novel isoforms. This repertoire provides invaluable insights for upcoming biological and theranostical studies for PDAC.

Chapter 5: This chapter summarized the whole thesis and provided potential directions for future research.