Page Comparison

Title: Understanding epigenetic changes in ovarian cancer

Project lead: Vitalina Komashko, Xia Yang

...

Complete description of the research plan with background and proposed methods can be found here. Also the latest brainstorming ideas about each specific aim of the project are outlined on the slides here.

Data access:

Data files and big files with results (everything that is not checked out to svn) can be found here: /work/DAT_002__TCGA_Ovarian/01_2011/vita_work

Table of contents:

1. Data description and collection
2. Normalization of DNA methylation data 
3. Analysis of clinical traits
   1. Survival
4. Analysis of gene expression data, smaller set of patients (~384)
   1. Correlation with clinical traits
      1. Survival
   2. And DNA methylation
      1. Correlation
   3. Gene coexpression network
      1. Analysis of chromosomal positions of genes within each module
      2. Correlation of coexpression PCs with DNA methylation loci
      3. Correlation of coexpression and comethylation PCs
      4. Gene overlap between modules of coexpression and comethylation networks
      5. Colocalization of coexpression and comethylation modules
5. Analysis of gene expression data, larger set of patients (522)
   1. Data retrieval and normalization
   2. Building a coexpression network
6. DNA methylation: correlation with clinical traits.
   1. Survival
7. DNA methylation network analysis (comethylation)
   1. Gene Ontology (GOcomethylation)
   2. Correlation of modules with survival (SurvivalComethylation)
   3. Analysis of chromosomal positions of CpGs within each module (CpGchromosome)
   4. Correlation of comethylation PCs with Affymetrix gene expression (smaller set of patients, ~384).
   5.  
8. DNA methylation analysis: expression methylation as M or beta value
   1. M value
      1. Univariate analysis
      2. Elastic net for the analysis of clinical traits 
      3. Network and network analysis
      4. Correlation of network modules with clinical traits
   2. Beta value
      1. Network and network analysis
      2. Correlation of network modules with clinical traits
   3. Summary and networks comparison

...

Summary of DNA methylation data analyses from TCGA ovarian cancer paper (PMID:21720365)

...

This project was discontinued in December 2011 because of the strong batch effect in DNA methylation data which was also highly correlated with clinical variables.