資料載入中.....
|
請使用永久網址來引用或連結此文件:
http://ir.cmu.edu.tw/ir/handle/310903500/10831
|
| 題名: | 整合性生物功能基因表現之對應分析
Correspondence Analysis of Gene Expression in Functional Groups |
| 作者: | 蔡政安(Tsai, Chen-An) |
| 貢獻者: | 公共衛生學院生物統計研究所 |
| 日期: | 2009-09-30 |
| 上傳時間: | 2009-09-01 14:52:09 (UTC+8) |
| 摘要: | 近年來，由於DNA微陣列生物晶片科技快速進展，提供研究人員得以結合各式各樣的臨床資料、實驗和生物功能註解來進行大規模全基因體的研究。然而，如何整合如此大量已知註解生物資訊和生物晶片資料來進行分析，已經成為一項非常嚴峻的挑戰。近來除了如何找出有顯著表現基因外，群體基因檢定(GCT)和群體基因富集分析(GSEA)也逐漸受到重視，認為是基因表現分析重要課題之一。ㄧ基因群體可能代表ㄧ群有相關功能的基因或者同屬於某一生物註解功能的類別；藉由加入這些基因的生物註解資訊進入基因表現分析，我們可以更深入去探討基因和基因及基因和生物特徵之間的交互作用分析。目前已有多種方法提出來分析在基因晶片實驗中，與代謝路徑或細胞調控結構有關之群體基因的顯著表現分析(例如: Mootha et al., 2003; Pavlidis et al., 2004; Tian et al., 2005)。然而，對於這些方法來檢測群體基因和生物特徵的相關性仍然未臻理想。本研究針對群體基因富集分析，利用整合已知生物註解功能的類別分類的基因群體，期望提出一合理實用的統計分析架構。在本計劃中我們將修正 O'Brien's OLS (1984) and Laüter's LWS (1996)的統計量來檢定有顯著表現改變的基因群體，期望能進一步提出能同時檢測up-regulated和down-regulated基因群體的統計量。此外，我們將利用生態學上常用之共慣性值分析(co-inertia analysis)來比較不同基因群的基因表現及相關性分析。在計畫中，各步驟的方法將透過電腦模擬的方式與其他方法比較。本研究主要結合基因晶片資料和在一些大型公用資料庫中已知的生物註解資訊來檢測及檢視基因群體表現之變化情形；期望能提供協助解釋複雜的生物演化過程。
Recent advances in DNA microarray technology have provided access to whole genome study combined with a variety of clinical outcomes, experimental and functional annotation, which has severely impacted on the investigation of array data. It is extremely challenging to integrate a certain expression profile of interest with diverse experimental settings and annotated biological processes. Recently, in addition to attempting to understand biological functions for each differentially expressed gene, gene-class testing (GCT) or gene set enrichment analysis (GSEA) have been proposed for gene expression analysis. A gene class refers to a group of genes with related functions or a set of genes grouped together based on biologically relevant information. The gene-gene and gene-phenotype interactions can be investigated by incorporating additional information into data analysis. In recent years various methods have been proposed to investigate differential expression profiles of groups of genes associated with the structure of pathways and cellular regulation (e.g., Mootha et al., 2003; Pavlidis et al., 2004; Tian et al., 2005) in microarray experiments. However, more fundamental criticisms raised by Damian et al. (2004) and Allison et al. (2006) refer to the capability of identifying relative strength of association with the phenotypes. In this project, a general statistical framework will be developed to provide a starting point of a complex and highly complicated process of interpretation. A fundamental property of biological systems is their ability to evolve, which depends on the underlying molecular mechanisms as well as on the relationships between groups of genes and the associated biological phenotypes. First, O'Brien's OLS (1984) and Laüter's LWS global statistics (1996) will be applied to testing significant alterations in the activities of gene sets for one-sided test, and a two-sided testing statistics will be proposed. Secondly, I will employ an approach called “co-inertia analysis”, which is a well-known approach for ecology, to compare expression profiles of gene sets. The co-inertia analysis (CIA) is a multivariate approach which is suitable for identifying trends or differences in expression of groups of genes which are functionally related. Such an analysis allows biologists to combine the extensive biological knowledge and annotations from public domain, such as Gene Ontology or KEGG, with microarray data for testing and visualizing changes in expression activity across groups of genes. |
| 顯示於類別: | [生物統計研究所] 研究計畫
|
文件中的檔案:
| 檔案 |
描述 |
大小 | 格式 | 瀏覽次數 |
|---|
| index.html | | 0Kb | HTML | 266 | 檢視/開啟 |
|
在CMUR中所有的資料項目都受到原著作權保護.
|
