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

Functional characterization of large...

Xie, Hongbo.

FindBook

Google Book

Amazon

博客來

Functional characterization of large scale biological data.

紀錄類型:	書目-語言資料,印刷品 : Monograph/item
正題名/作者:	Functional characterization of large scale biological data./
作者:	Xie, Hongbo.
面頁冊數:	261 p.
附註:	Adviser: Zoran Obradovic.
Contained By:	Dissertation Abstracts International69-01B.
標題:	Biology, Bioinformatics. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3300385
ISBN:	9780549447139

Functional characterization of large scale biological data.
Xie, Hongbo.

Functional characterization of large scale biological data. - 261 p.

Adviser: Zoran Obradovic.

Thesis (Ph.D.)--Temple University, 2008.

Rapid growth in collection of new gene and protein related information requires development of appropriate data analysis techniques. Towards such a goal in this study we propose incorporating prior knowledge of biological functions of certain proteins into early stages of data analysis process. As the first aim of our study, a novel method aimed at functionally characterizing gene patterns over time is proposed. Our approach is based on analysis of functional expression profiles (FEP), each defined as the average of expression patterns for genes annotated with a given function. A FEP is computed only over significantly correlated patterns that also vary significantly in time. In our study an effective clustering method is proposed to automatically discover the most informative groups of FEPs. The new method is evaluated on several important time-course gene expression datasets (including development cycle of malaria-related Plasmodium Falciparum) where it successfully identified correlated functional expression profiles. Furthermore, clustering these functional expression profiles provided groups of functions with similar expression profile pattern and close biological functional relevance. These results indicate that the analysis method could lead to novel biological conclusions and benefit research on various types of Microarray data. As the second aim, we proposed a novel technique for exploiting knowledge of a biological function into biomarker candidates' identification. In our two-step approach for selection of genetic biomarkers from microarray data the underlying assumption is that disease is characterized by deviations in expression of genes from a limited set of functions. We start by selection of significantly differentially expressed genes by using a standard statistical testing procedure. Using functional domain knowledge, we analyze biological functions of the selected genes to discover the ones that are highly overrepresented by the selection. Only the selected genes annotated with the most significant function are selected as biomarker candidates. The new method is applied for identification of biomarkers for Chronic Fatigue Syndrome (CFS). The approach resulted in a small set of biomarkers whose functions are the most relevant to CFS that was superior to a much larger set determined the traditional one-step analysis. We also explored benefits of combining Microarray and proteomics data for CFS identification. Using the standard procedure for preprocessing of ProteinChip data, we developed a proteomics-based predictor of CFS. The results on the samples with both Microarray and ProteinChip data indicates that predictor combination can provide improved CFS identification. Our analysis of the clinical CFS data identified factors that explain sources of CFS identification mistakes suggesting that CSF identification could be further improved by revising definitions of certain clinical conditions. In the last aim we developed a bioinformatics method to identify structurally related functions in disordered proteins. In our method a statistical evaluation is employed to rank the significance of identified correlations where protein sequence data redundancy and the relationship between protein length and protein structure were taken into consideration to ensure the quality of the statistical inference. We applied the new method on Swiss-Prot database to identify intrinsic disorder correlated functional keywords. This work enriches the current view of protein structure-function relationships, especially with regards to functionalities of intrinsically disordered proteins and provides researchers with a novel tool that could be used to improve the understanding of the relationship between protein structure and function. (Abstract shortened by UMI.)

ISBN: 9780549447139Subjects--Topical Terms:

1018415
Biology, Bioinformatics.

Functional characterization of large scale biological data.
LDR:04681nam 2200277 a 45 001 963023
005 20110830
008 110831s2008 ||||||||||||||||| ||eng d
020 $a 9780549447139
035 $a (UMI)AAI3300385
035 $a AAI3300385
040 $a UMI $c UMI
100 1 $a Xie, Hongbo. $3 1286082
245 1 0 $a Functional characterization of large scale biological data.
300 $a 261 p.
500 $a Adviser: Zoran Obradovic.
500 $a Source: Dissertation Abstracts International, Volume: 69-01, Section: B, page: 0444.
502 $a Thesis (Ph.D.)--Temple University, 2008.
520 $a Rapid growth in collection of new gene and protein related information requires development of appropriate data analysis techniques. Towards such a goal in this study we propose incorporating prior knowledge of biological functions of certain proteins into early stages of data analysis process. As the first aim of our study, a novel method aimed at functionally characterizing gene patterns over time is proposed. Our approach is based on analysis of functional expression profiles (FEP), each defined as the average of expression patterns for genes annotated with a given function. A FEP is computed only over significantly correlated patterns that also vary significantly in time. In our study an effective clustering method is proposed to automatically discover the most informative groups of FEPs. The new method is evaluated on several important time-course gene expression datasets (including development cycle of malaria-related Plasmodium Falciparum) where it successfully identified correlated functional expression profiles. Furthermore, clustering these functional expression profiles provided groups of functions with similar expression profile pattern and close biological functional relevance. These results indicate that the analysis method could lead to novel biological conclusions and benefit research on various types of Microarray data. As the second aim, we proposed a novel technique for exploiting knowledge of a biological function into biomarker candidates' identification. In our two-step approach for selection of genetic biomarkers from microarray data the underlying assumption is that disease is characterized by deviations in expression of genes from a limited set of functions. We start by selection of significantly differentially expressed genes by using a standard statistical testing procedure. Using functional domain knowledge, we analyze biological functions of the selected genes to discover the ones that are highly overrepresented by the selection. Only the selected genes annotated with the most significant function are selected as biomarker candidates. The new method is applied for identification of biomarkers for Chronic Fatigue Syndrome (CFS). The approach resulted in a small set of biomarkers whose functions are the most relevant to CFS that was superior to a much larger set determined the traditional one-step analysis. We also explored benefits of combining Microarray and proteomics data for CFS identification. Using the standard procedure for preprocessing of ProteinChip data, we developed a proteomics-based predictor of CFS. The results on the samples with both Microarray and ProteinChip data indicates that predictor combination can provide improved CFS identification. Our analysis of the clinical CFS data identified factors that explain sources of CFS identification mistakes suggesting that CSF identification could be further improved by revising definitions of certain clinical conditions. In the last aim we developed a bioinformatics method to identify structurally related functions in disordered proteins. In our method a statistical evaluation is employed to rank the significance of identified correlations where protein sequence data redundancy and the relationship between protein length and protein structure were taken into consideration to ensure the quality of the statistical inference. We applied the new method on Swiss-Prot database to identify intrinsic disorder correlated functional keywords. This work enriches the current view of protein structure-function relationships, especially with regards to functionalities of intrinsically disordered proteins and provides researchers with a novel tool that could be used to improve the understanding of the relationship between protein structure and function. (Abstract shortened by UMI.)
590 $a School code: 0225.
650 4 $a Biology, Bioinformatics. $3 1018415
650 4 $a Computer Science. $3 626642
690 $a 0715
690 $a 0984
710 2 $a Temple University. $3 959342
773 0 $t Dissertation Abstracts International $g 69-01B.
790 $a 0225
790 1 0 $a Obradovic, Zoran, $e advisor
791 $a Ph.D.
792 $a 2008
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3300385