Please use this identifier to cite or link to this item:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/89762
Title: | 用高通量基因資料建構不同癌症類型病患的預後指標 Development of Prognostic Models for Patients with Different Cancer Types by Using High-throughput Genomic Data |
Authors: | 蕭亦文 Yi-Wen Hsiao |
Advisor: | 盧子彬 Tzu-Pin Lu |
Keyword: | 基因表現量,腫瘤浸潤淋巴細胞,同源重組修復缺失,卵巢癌,肝癌,泛癌症研究,競爭型核糖核酸, Gene expression,Tumor-infiltrating lymphocytes,Homologous recombination deficiency,Ovarian cancer,Hepatocellular carcinomas,Pan-cancer study,Competing endogenous RNA, |
Publication Year : | 2023 |
Degree: | 博士 |
Abstract: | 隨著高通量基因定量技術的普及以及生物資訊相關演算法的日新月異,使得大型基因體數據的取得與深入分析不再遙不可及。雖然臨床資料具有取得性以及 預後預測準確性的優勢,但仍有部分的預後預測是需要透過基因資訊來進一步解釋。所以如何有效地找尋適合不同癌症類型的生物標記來作為檢測或評估癌症疾病的發展進而設計客製化的治療方針是目前生物醫學領域的一大課題。故本論文旨在不同的基因層級上找出針對不同癌症類型病患的預後指標。本論文分成四部分: (1)比較同源重組修復缺失的分數和相對應的預後情況在非裔、歐裔和亞裔美國人之間的差異,再進一步探討哪種基因特徵類型以及那個基因集可以有效的檢測同源重組修復缺失狀態; (2)透過機器學習的方式找出有效的基因表現量圖譜來建構一個可以高靈敏度地找出高風險的卵巢癌患者風險預測模型; (3)比較 B 型肝癌和 C 型肝癌患者之間所有腫瘤浸潤淋巴細胞的差異以及其預後的情況; (4)欲建構一個可以有效找出癌症相關的競爭型核糖核酸的演算法。分別得到以下的結果: (1)找出族群特異性的同源重組修復缺失基因突變圖譜,有助於客製化建立各個族群的篩選源重組修復缺失現象病患進而提供特定的治療方針; (2)建構一個高靈敏度的高死亡風險卵巢癌患者預測模型,且選用的基因集特徵需與卵巢癌的生物機制有相關; (3)了解在腫瘤與非腫瘤組織的腫瘤浸潤淋巴細胞的總量差異在不同的肝癌亞型中是否一致,以及找出在不同肝癌亞型中哪些免疫細胞是跟預後相關的生物標記指標; (4) 建構一個穩健且可應用的找尋競爭型核糖核酸的演算法,並透過一些下游分析來進一步解讀這樣的基因調控事件。所以透過這些研究來找尋癌症特有的基因特徵,進而了解在癌症種類或是其亞型與族群間的預後差異,可以針對不同癌症或是其亞型與族群設計客製化檢測工具,以達到精準醫療的願景。 With the rapid development of the next-generation sequencing techniques and bioinformatics algorithms, high-throughput biological data have become less expensive and more accessible such that more studies can explore to a greater level the genetic impacts on various diseases. Although clinical data has the advantage of its accessibility and accuracy of prognostic prediction, 30-40% of patients fail to use clinical factors to evaluate their prognostic outcomes. Therefore, effectively identifying genomic biomarkers for the prognostic evaluation and the guidance of the necessary medical intervention remains the main challenge in the biomedical field. Hence, this dissertation aims to define prognostic biomarkers specifically for patients with particular cancer types using DNA- or RNA-level genomic data. There are three specific objectives in this dissertation: (1) to elucidate the racial differences in HRD scar scores in multiple cancers and to evaluate their associations with clinical outcomes; also, to assess the applicability of each HRD gene-sets for each cancer-race group. (2) to develop a bagging-based algorithm with GA-XGBoost models that predicts the high risk of death from ovarian cancer using gene expression profiles; (3) to define the cell composition of the immune response in both HBV−HCC and HCV−HCC and to investigate its relationship with clinical outcomes such as overall survival and recurrence-free survival; (4) to effectively identify the massive number of known ceRNA interactions using genome-wide transcriptome and miRNA profiles and to interpret their associations with cancers. Through the above studies, the results revealed that (1) a race-specific profile of the predisposing HRD-related genes with mutations was identified for customized design of HRD screening approaches. (2) a robust risk prediction model using selected gene expression related to ovarian cancer with a high risk of death will be constructed; (3) whether the difference of the total amount of TILs in tumor and non-tumor tissues is consistent in different liver cancer subtypes, and the biomarkers that associated with survival outcomes between two subtypes of hepatocellular carcinomas will be unveiled; (4) a computational framework for the identification of ceRNA events and their biological interpretations. In conclusion, through these studies, we would be able to identify specific genomic features of cancers, understand the prognostic difference among cancer subtypes or racial groups, and design the customized tumor testing tools for specific cancer types or racial groups, reaching the ultimate goal of precision medicine. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/89762 |
DOI: | 10.6342/NTU202300922 |
Fulltext Rights: | 同意授權(全球公開) |
Appears in Collections: | 流行病學與預防醫學研究所 |
Files in This Item:
File | Size | Format | |
---|---|---|---|
ntu-111-2.pdf | 6.41 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.