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

Selection of Disease-Associated Gene...

Kim, Hoon.

FindBook

Google Book

Amazon

博客來

Selection of Disease-Associated Gene Sets.

紀錄類型:	書目-語言資料,印刷品 : Monograph/item
正題名/作者:	Selection of Disease-Associated Gene Sets./
作者:	Kim, Hoon.
面頁冊數:	144 p.
附註:	Source: Dissertation Abstracts International, Volume: 72-06, Section: B, page: .
Contained By:	Dissertation Abstracts International72-06B.
標題:	Biology, Genetics. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3451816
ISBN:	9781124572628

Selection of Disease-Associated Gene Sets.
Kim, Hoon.

Selection of Disease-Associated Gene Sets. - 144 p.

Source: Dissertation Abstracts International, Volume: 72-06, Section: B, page: .

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

Recent advances in DNA microarray technology have given researchers an incredible opportunity to better understand disease. High-throughput technology has been already used to identify new molecular subtypes of disease and discover effective therapeutic targets. In order to extract clinically useful information from massive amounts of biological data, novel computational techniques have been needed. Our research presented in this thesis fills this void by proposing computational approaches to select gene sets for 1) classification and 2) molecular subtyping, so as to better understand the biological mechanisms associated with disease and, thus, improve disease treatments.

ISBN: 9781124572628Subjects--Topical Terms:

1017730
Biology, Genetics.

Selection of Disease-Associated Gene Sets.
LDR:04773nam 2200301 4500 001 1403042
005 20111108080400.5
008 130515s2011 ||||||||||||||||| ||eng d
020 $a 9781124572628
035 $a (UMI)AAI3451816
035 $a AAI3451816
040 $a UMI $c UMI
100 1 $a Kim, Hoon. $3 1254390
245 1 0 $a Selection of Disease-Associated Gene Sets.
300 $a 144 p.
500 $a Source: Dissertation Abstracts International, Volume: 72-06, Section: B, page: .
502 $a Thesis (Ph.D.)--Columbia University, 2011.
520 $a Recent advances in DNA microarray technology have given researchers an incredible opportunity to better understand disease. High-throughput technology has been already used to identify new molecular subtypes of disease and discover effective therapeutic targets. In order to extract clinically useful information from massive amounts of biological data, novel computational techniques have been needed. Our research presented in this thesis fills this void by proposing computational approaches to select gene sets for 1) classification and 2) molecular subtyping, so as to better understand the biological mechanisms associated with disease and, thus, improve disease treatments.
520 $a Regarding gene selection for classification, we propose two methods. One, which is named "synergistic recursive feature elimination," attempts to use the complementary strengths of two existing gene selection approaches, which are information- maximization and recursive feature elimination, by retaining an optimum small subset of features by maximizing the information that it provides about the class label and then augmenting this small subset by selecting another, larger, set of genes, serving as its "synergistic partner," using recursive feature elimination. We demonstrate that its improved performance, compared with traditional recursive feature elimination, is not due to the inclusion of the small maximum-information subset to the final set of biomarkers; rather, it is due to the synergistic association of that small subset with the set of remaining genes, with respect to the class label.
520 $a The second approach of gene selection for classification we propose is the use of statistical tests in biomarker discovery. To reduce the risk of selecting spurious gene subsets, a common problem when using multivariate filter methods for selecting features, the proposed approach employs two types of significance tests for selecting small gene subsets, such as pairs that are associated with class labels: an "overall significance test" and an "incremental significance test." The overall significance test examines whether or not, say, a gene pair is selected by pure chance. The incremental significance test compares the correlation of the pair with that of the most discriminatory gene (the "main" gene) between the two member genes, examining whether or not the observed improvement following the addition of the "partner" gene is due to pure chance. The proposed approach is capable of identifying gene sets, each of which is generated by significant interaction between its members. Furthermore, compared to established approaches, this method offers enhanced classification performance.
520 $a In addition to gene selection for classification, we provide a novel gene selection method for class discovery identifying sets of genes whose coordinated expression indicates the presence of a particular phenotype. Analyzing data from multiple cancers, our proposed approach reveals a core multi-cancer invasion-associated gene signature that is triggered at particular stages of cancers. The identified genes suggest that a key feature of the gene signature is a special type of TGF-beta signaling-induced by a ligand known as activin A. Furthermore, this work establishes that the core signature we identify can predict neoadjuvant therapy response in node-negative breast cancer. Our findings from this work can be used for developing high specificity biomarkers sensing cancer invasion, and developing potential multi- cancer metastasis-inhibiting therapeutics that target the corresponding biological mechanism.
520 $a In the near future, a vast amount of additional biological information will become available, including next generation sequencing, miRNA and DNA methylation for many cancer types, which will allow additional computational research to build on this research and clarify the details of the underlying cancer-associated complex biological process.
590 $a School code: 0054.
650 4 $a Biology, Genetics. $3 1017730
650 4 $a Biology, Bioinformatics. $3 1018415
690 $a 0369
690 $a 0715
710 2 $a Columbia University. $3 571054
773 0 $t Dissertation Abstracts International $g 72-06B.
790 $a 0054
791 $a Ph.D.
792 $a 2011
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3451816