東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Integrative computational analysis, ...

Boston University.

FindBook

Google Book

Amazon

博客來

Integrative computational analysis, design and validation of cancer biomarkers via functional modules and pheno-genomic profiling.

紀錄類型:	書目-語言資料,印刷品 : Monograph/item
正題名/作者:	Integrative computational analysis, design and validation of cancer biomarkers via functional modules and pheno-genomic profiling./
作者:	Wu, Chang-Jiun.
面頁冊數:	189 p.
附註:	Adviser: Simon Kasif.
Contained By:	Dissertation Abstracts International70-01B.
標題:	Biology, Genetics. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoeng/servlet/advanced?query=3345601
ISBN:	9781109001884

Integrative computational analysis, design and validation of cancer biomarkers via functional modules and pheno-genomic profiling.
Wu, Chang-Jiun.

Integrative computational analysis, design and validation of cancer biomarkers via functional modules and pheno-genomic profiling. - 189 p.

Adviser: Simon Kasif.

Thesis (Ph.D.)--Boston University, 2009.

These results, though preliminary, show great promise and provide potential opportunities to search for better clinical or non-invasive biomarkers in various diseases.

ISBN: 9781109001884Subjects--Topical Terms:

1017730
Biology, Genetics.

Integrative computational analysis, design and validation of cancer biomarkers via functional modules and pheno-genomic profiling.
LDR:03680nam 2200325 a 45 001 856987
005 20100709
008 100709s2009 ||||||||||||||||| ||eng d
020 $a 9781109001884
035 $a (UMI)AAI3345601
035 $a AAI3345601
040 $a UMI $c UMI
100 1 $a Wu, Chang-Jiun. $3 1023886
245 1 0 $a Integrative computational analysis, design and validation of cancer biomarkers via functional modules and pheno-genomic profiling.
300 $a 189 p.
500 $a Adviser: Simon Kasif.
500 $a Source: Dissertation Abstracts International, Volume: 70-01, Section: B, page: 0467.
502 $a Thesis (Ph.D.)--Boston University, 2009.
520 $a These results, though preliminary, show great promise and provide potential opportunities to search for better clinical or non-invasive biomarkers in various diseases.
520 $a Cancer is a heterogeneous disease associated with complex multigenic mechanisms and multistep processes. Connections between phenotypic variations and molecular signatures obtained via genome-wide technologies advance our understanding of tumor biology and can lead towards improved diagnostic, and therapeutic strategies. The first part of this thesis introduces a contextually novel methodology that enables rigorous comparison of molecular and clinical cancer biomarkers across multiple patient cohorts. We critically evaluated breast cancer prognostic biomarkers including precursors of commercially available signatures, conventional clinical models, and genomic signatures designed to track basic tumor properties. Our analysis indicates that all genomic signatures predict outcomes concordantly across a broad set of studies. Existing biomarkers appear to track primarily the relationship of cell cycle activity to prognosis. The performance of clinical models is on par with the top genomic biomarkers.
520 $a Biomarkers aimed to discover basic tumor properties other than proliferative activities have the potential to improve outcome prediction. The second part of the thesis describes a Pheno-Genomic Module Profiler (PGMP) algorithm that seeks to discover groups of functionally constrained, co-expressed, transcriptionally co-regulated genes controlling phenotype-associated processes. By integrating microarray expression profiles, systematic pathway annotations, transcription factor binding information, and protein-interaction network data, we detected three independent prognostic gene modules robustly predictive of cancer outcome in multiple breast cancer cohorts. They include two pro-tumor modules involved with cell cycle activity and extracellular matrix interaction, and an anti-tumor module enriched with genes from multiple immune-related pathways. A model combining these modules predicts outcomes better than signatures based solely on cell cycle activities. In the estrogen receptor-negative breast cancer patients, the new model improves the predictive performance by 10&sim;20% compared to existing biomarkers.
520 $a Motivated by the fact that transcriptional profiling is difficult to obtain in clinical situations, the third part of this thesis develops a Genomic Information Content (GIC) method. This new model amplifies the predictive values of clinical variables via a projection into signature genomic states, and shows a slightly better performance than conventional models in cross validation.
590 $a School code: 0017.
650 4 $a Biology, Genetics. $3 1017730
650 4 $a Biology, Molecular. $3 1017719
650 4 $a Engineering, Biomedical. $3 1017684
690 $a 0307
690 $a 0369
690 $a 0541
710 2 $a Boston University. $3 1017454
773 0 $t Dissertation Abstracts International $g 70-01B.
790 $a 0017
790 1 0 $a Kasif, Simon, $e advisor
791 $a Ph.D.
792 $a 2009
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoeng/servlet/advanced?query=3345601