國立虎尾科技大學 |

Analysis of Aberrant Regulation of Gene Expression in Cancer.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Analysis of Aberrant Regulation of Gene Expression in Cancer./
作者:	Wu, Pamela.
面頁冊數:	1 online resource (139 pages)
附註:	Source: Dissertation Abstracts International, Volume: 79-08(E), Section: B.
Contained By:	Dissertation Abstracts International79-08B(E).
標題:	Bioinformatics. -
電子資源:	click for full text (PQDT)
ISBN:	9780355773613

Analysis of Aberrant Regulation of Gene Expression in Cancer.
Wu, Pamela.

Analysis of Aberrant Regulation of Gene Expression in Cancer. - 1 online resource (139 pages)

Source: Dissertation Abstracts International, Volume: 79-08(E), Section: B.

Thesis (Ph.D.)--New York University, 2018.

Includes bibliographical references

The intersection of developments in cancer biology, high-throughput molecular assays, and machine learning has created a vast array of new questions and challenges on both biological and computational fronts. In the first chapter, we deconstruct the current way signal values are used as features in machine learning by creating orthogonal features representing combinations of signal values in a attempt to see if model accuracy, biological generalizability, and/or stability of information between models is improved by either feature set. This concept was applied to histone modification signal values and their significant combinations, chromatin states, for the prediction of gene expression, coding vs. lncRNA, and cell-type specificity values at gene loci because histone modification patterns at loci has been shown to be strongly associated with gene regulation. We found that for both model accuracy and biological generalizability, gene expression prediction was best served by signal value features and coding vs. lncRNA was best served by chromatin states features. Chromatin states features were consistently more likely to be selected during feature selection and also showed a strong ability to preserve histone modification importance rankings between linear and non-linear models. The next two chapters describe applications and development of methods to analyse cancer genomics data. The first study describes the differential expression analysis performed to find candidates for a loss-of-function screen that identified AMIGO2 as a melanoma survival gene, followed by analysis of transcription factor motifs, histone modification signal maps, and chromatin states to explore its epigenetic context. Next, in order to examine the protein composition of extracellular matrix structures involved in non-endothelial vascularization in optic gliomas, a method for RNA-seq differential expression analysis was adapted for mass spectrometry spectral counts and the resuls were used to build a protein-protein interaction graph with overlaid expression data. This method identifies clusters of significantly expressed genes or proteins, which can guide research into novel physiological structures. Lastly, one challenge of the increasing volume of omics data is the question of where to store the data while exposing it in a way that allows for easy integrative analysis and data exploration. In the last chapter, mass spectrometry data from selected The Cancer Genome Atlas samples that were assayed for protein composition via mass spectrometry were added to the cBioPortal interface, a web application that facilitates exploration and visualization of cancer genomics data. Scripts to transform data to be ingested by cBioPortal were made to support both TCGA and MaxQuant format files, with an option to use the proteomic ruler method that converts mass spectrometry signal values into absolute protein copy number per cell. An additional heatmap component was created to complement the new data.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2018

Mode of access: World Wide Web

ISBN: 9780355773613Subjects--Topical Terms:

583857
Bioinformatics.
Index Terms--Genre/Form:

554714
Electronic books.

Analysis of Aberrant Regulation of Gene Expression in Cancer.
LDR:04219ntm a2200337Ki 4500 001 919123
005 20181116131020.5
006 m o u
007 cr mn||||a|a||
008 190606s2018 xx obm 000 0 eng d
020 $a 9780355773613
035 $a (MiAaPQ)AAI10681951
035 $a (MiAaPQ)nyu:13131
035 $a AAI10681951
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Wu, Pamela. $3 1193624
245 1 0 $a Analysis of Aberrant Regulation of Gene Expression in Cancer.
264 0 $c 2018
300 $a 1 online resource (139 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertation Abstracts International, Volume: 79-08(E), Section: B.
500 $a Adviser: David Fenyo.
502 $a Thesis (Ph.D.)--New York University, 2018.
504 $a Includes bibliographical references
520 $a The intersection of developments in cancer biology, high-throughput molecular assays, and machine learning has created a vast array of new questions and challenges on both biological and computational fronts. In the first chapter, we deconstruct the current way signal values are used as features in machine learning by creating orthogonal features representing combinations of signal values in a attempt to see if model accuracy, biological generalizability, and/or stability of information between models is improved by either feature set. This concept was applied to histone modification signal values and their significant combinations, chromatin states, for the prediction of gene expression, coding vs. lncRNA, and cell-type specificity values at gene loci because histone modification patterns at loci has been shown to be strongly associated with gene regulation. We found that for both model accuracy and biological generalizability, gene expression prediction was best served by signal value features and coding vs. lncRNA was best served by chromatin states features. Chromatin states features were consistently more likely to be selected during feature selection and also showed a strong ability to preserve histone modification importance rankings between linear and non-linear models. The next two chapters describe applications and development of methods to analyse cancer genomics data. The first study describes the differential expression analysis performed to find candidates for a loss-of-function screen that identified AMIGO2 as a melanoma survival gene, followed by analysis of transcription factor motifs, histone modification signal maps, and chromatin states to explore its epigenetic context. Next, in order to examine the protein composition of extracellular matrix structures involved in non-endothelial vascularization in optic gliomas, a method for RNA-seq differential expression analysis was adapted for mass spectrometry spectral counts and the resuls were used to build a protein-protein interaction graph with overlaid expression data. This method identifies clusters of significantly expressed genes or proteins, which can guide research into novel physiological structures. Lastly, one challenge of the increasing volume of omics data is the question of where to store the data while exposing it in a way that allows for easy integrative analysis and data exploration. In the last chapter, mass spectrometry data from selected The Cancer Genome Atlas samples that were assayed for protein composition via mass spectrometry were added to the cBioPortal interface, a web application that facilitates exploration and visualization of cancer genomics data. Scripts to transform data to be ingested by cBioPortal were made to support both TCGA and MaxQuant format files, with an option to use the proteomic ruler method that converts mass spectrometry signal values into absolute protein copy number per cell. An additional heatmap component was created to complement the new data.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2018
538 $a Mode of access: World Wide Web
650 4 $a Bioinformatics. $3 583857
650 4 $a Oncology. $3 593951
655 7 $a Electronic books. $2 local $3 554714
690 $a 0715
690 $a 0992
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a New York University. $b Biology. $3 1193625
773 0 $t Dissertation Abstracts International $g 79-08B(E).
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10681951 $z click for full text (PQDT)