國立高雄大學圖資館 |

Language: English

Back

Identifying functional modules from ...

Boston University.

Identifying functional modules from biological networks and expression profiles.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Identifying functional modules from biological networks and expression profiles.
Author:	Hung, Jui-hung.
Description:	129 p.
Notes:	Source: Dissertation Abstracts International, Volume: 72-09, Section: B, page: 5045.
Notes:	Adviser: Charles DeLisi.
Contained By:	Dissertation Abstracts International72-09B.
Subject:	Biology, Systematic.
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3463153
ISBN:	9781124755151

Identifying functional modules from biological networks and expression profiles.
Hung, Jui-hung.

Identifying functional modules from biological networks and expression profiles. - 129 p.

Source: Dissertation Abstracts International, Volume: 72-09, Section: B, page: 5045.

Thesis (Ph.D.)--Boston University, 2011.

One of the major goals of systems biology is to understand functional modules, or the mechanisms that carry out life-sustaining biological functions at the molecular level. The term 'functional modules' encompasses not only sets of genes or proteins, but also the interactions between them. This dissertation undertakes the following two strategies to identify these functional cornerstones: (1) First, by comparing the distinct collective behaviors (i.e. expressions) of a priori defined and functional related gene sets of different phenotypes. This method of identifying the causal network modules driving the phenotypic behaviors (rather than an individual gene) is becoming increasingly popular in the field. Currently, Enrichment Analysis (EA) is one of the most widely applied statistical approaches for this task. In this dissertation I propose a novel improvement to EA by introducing gene interactions into a weighted statistical procedure, which yields better specificity and sensitivity in discovering cancer-related pathways. In addition, I review and evaluate a wide range of EA variants based on a novel mutual coverage method when the gold standard is absent. (2) Second, by finding conserved modules of interaction network across different species. Crucial functional interaction modules (i.e. pathways) are obtained through eons of evolution. Therefore, the identification of conserved interaction sub-networks within multiple species is an informative resource in understanding biological mechanisms. Network alignment algorithms were proposed to find these conserved modules, but, due to the complexity of the problem, current methods are incapable of dealing with large-scale data. I transform the problem to a metagraph and apply slime mold modeling (SMM) to identify conserved modules, and visualize them in an intuitive representation. In examining the five most conserved predicted modules, I find that four of them overlap with known pathways significantly. The one that does not overlap likely represents a newly documented tmRNA mechanism.

ISBN: 9781124755151Subjects--Topical Terms:

603175
Biology, Systematic.

Identifying functional modules from biological networks and expression profiles.
LDR:03361nmm 2200301 4500 001 380614
005 20130530092659.5
008 130708s2011 ||||||||||||||||| ||eng d
020 $a 9781124755151
035 $a (UMI)AAI3463153
035 $a AAI3463153
040 $a UMI $c UMI
100 1 $a Hung, Jui-hung. $3 603174
245 1 0 $a Identifying functional modules from biological networks and expression profiles.
300 $a 129 p.
500 $a Source: Dissertation Abstracts International, Volume: 72-09, Section: B, page: 5045.
500 $a Adviser: Charles DeLisi.
502 $a Thesis (Ph.D.)--Boston University, 2011.
520 $a One of the major goals of systems biology is to understand functional modules, or the mechanisms that carry out life-sustaining biological functions at the molecular level. The term 'functional modules' encompasses not only sets of genes or proteins, but also the interactions between them. This dissertation undertakes the following two strategies to identify these functional cornerstones: (1) First, by comparing the distinct collective behaviors (i.e. expressions) of a priori defined and functional related gene sets of different phenotypes. This method of identifying the causal network modules driving the phenotypic behaviors (rather than an individual gene) is becoming increasingly popular in the field. Currently, Enrichment Analysis (EA) is one of the most widely applied statistical approaches for this task. In this dissertation I propose a novel improvement to EA by introducing gene interactions into a weighted statistical procedure, which yields better specificity and sensitivity in discovering cancer-related pathways. In addition, I review and evaluate a wide range of EA variants based on a novel mutual coverage method when the gold standard is absent. (2) Second, by finding conserved modules of interaction network across different species. Crucial functional interaction modules (i.e. pathways) are obtained through eons of evolution. Therefore, the identification of conserved interaction sub-networks within multiple species is an informative resource in understanding biological mechanisms. Network alignment algorithms were proposed to find these conserved modules, but, due to the complexity of the problem, current methods are incapable of dealing with large-scale data. I transform the problem to a metagraph and apply slime mold modeling (SMM) to identify conserved modules, and visualize them in an intuitive representation. In examining the five most conserved predicted modules, I find that four of them overlap with known pathways significantly. The one that does not overlap likely represents a newly documented tmRNA mechanism.
520 $a To sum up, this dissertation proposes improved methodologies that utilize the interactions between genes and proteins in order to facilitate the discovery of functional modules, offering insight into biological functions. The dissertation includes an in-depth discussion of case studies to illustrate the usefulness of my algorithms when applied to real biological data.
590 $a School code: 0017.
650 4 $a Biology, Systematic. $3 603175
650 4 $a Engineering, Biomedical. $3 227004
650 4 $a Biology, Bioinformatics. $3 264207
690 $a 0423
690 $a 0541
690 $a 0715
710 2 $a Boston University. $3 212722
773 0 $t Dissertation Abstracts International $g 72-09B.
790 1 0 $a DeLisi, Charles, $e advisor
790 $a 0017
791 $a Ph.D.
792 $a 2011
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3463153