東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Mining gene-phenotype networks using...

Columbia University.

Linked to FindBook

Google Book

Amazon

博客來

Mining gene-phenotype networks using semantic similarity.

Record Type:	Language materials, printed : Monograph/item
Title/Author:	Mining gene-phenotype networks using semantic similarity./
Author:	Tao, Ying.
Description:	202 p.
Notes:	Adviser: Yves A. Lussier.
Contained By:	Dissertation Abstracts International68-01B.
Subject:	Biology, Bioinformatics. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3249135

Mining gene-phenotype networks using semantic similarity.
Tao, Ying.

Mining gene-phenotype networks using semantic similarity. - 202 p.

Adviser: Yves A. Lussier.

Thesis (Ph.D.)--Columbia University, 2007.

The discovery of relationships between genotypes and their emergent phenotypes is a central focus of modern biology. Taken together, these gene-phenotype and phenotype-phenotype associative relationships, along with the ontological organization of phenotypes, constitute Gene-Phenotype (GP) networks. A number of significant challenges stand in the way of mining such networks, including the heterogeneity of phenotype information, its hierarchical organization in biomedical ontologies, and its rapidly expanding volume. In this research, I hypothesize that an information theory-based semantic similarity algorithm, in conjunction with biological reference ontologies, can be used to predict new and valid gene-phenotype associations when applied to complex GP networks constructed from these ontologies.Subjects--Topical Terms:

1018415
Biology, Bioinformatics.

Mining gene-phenotype networks using semantic similarity.
LDR:03260nam 2200265 a 45 001 852646
005 20100630
008 100630s2007 ||||||||||||||||| ||eng d
035 $a (UMI)AAI3249135
035 $a AAI3249135
040 $a UMI $c UMI
100 1 $a Tao, Ying. $3 1018621
245 1 0 $a Mining gene-phenotype networks using semantic similarity.
300 $a 202 p.
500 $a Adviser: Yves A. Lussier.
500 $a Source: Dissertation Abstracts International, Volume: 68-01, Section: B, page: 0028.
502 $a Thesis (Ph.D.)--Columbia University, 2007.
520 $a The discovery of relationships between genotypes and their emergent phenotypes is a central focus of modern biology. Taken together, these gene-phenotype and phenotype-phenotype associative relationships, along with the ontological organization of phenotypes, constitute Gene-Phenotype (GP) networks. A number of significant challenges stand in the way of mining such networks, including the heterogeneity of phenotype information, its hierarchical organization in biomedical ontologies, and its rapidly expanding volume. In this research, I hypothesize that an information theory-based semantic similarity algorithm, in conjunction with biological reference ontologies, can be used to predict new and valid gene-phenotype associations when applied to complex GP networks constructed from these ontologies.
520 $a In this research, I developed a novel approach, Information Theory-based Semantic Similarity (ITSS), to automatically predict new phenotypic annotations of genes based on their known Gene Ontology (GO) annotations. The use of semantic similarity measurement based on information theory in GO allows for the representation of ontologically organized knowledge as mathematically-defined values, which address the problems of heterogeneous granularity of phenotypes. I demonstrate, using a 10-fold cross-validation, that the ITSS algorithm obtains prediction accuracies (Precision 97%, Recall 77%) comparable to other machine learning algorithms when applied to similar densely annotated portions of the GO annotations. In addition, such a method can generate highly accurate predictions in sparsely annotated portions of GO annotations, in which previous algorithm failed to do so. As a result, the technique generates more gene function predictions than previous methods. Furthermore, this study presents the first historical rollback validation for the predicted GO annotations, which may represent more realistic conditions for an evaluation than the generally used cross-validation approach. The use of semantic relations between concepts can increase the prediction accuracy by up to 12.7% in predicting future gene annotations as demonstrated in the historical roll-back validation. By manually assessing a random sample of 100 predictions conducted in a historical roll-back evaluation, we estimate that a minimum precision of 51% (95% confidence interval: 43%--58%) can be achieved for a human GO annotation file dated 2003.
590 $a School code: 0054.
650 4 $a Biology, Bioinformatics. $3 1018415
690 $a 0715
710 2 $a Columbia University. $3 571054
773 0 $t Dissertation Abstracts International $g 68-01B.
790 $a 0054
790 1 0 $a Lussier, Yves A., $e advisor
791 $a Ph.D.
792 $a 2007
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3249135