Mammary Cancer and Activation of Transposable Elements
Abstract
We have made significant methodological improvements to streamline our Methyl-MAPS protocol such that it uses less input DNA while substantially reducing processing time. We also developed a new computational algorithm that utilizes the entire methylation profile in the vicinity of each gene promoter to discover patterns of methylation changes that correlate with transcription. Methylation data is conventionally analyzed using sliding windows to identify regions of differential methylation, which are at best weakly correlated with expression of nearby genes. Since annotated promoters produce only weak correlations, applying such tools to understand gene regulation by demethylation of transposable elements would be difficult. Application of our method shows that when we consider the entire methylation profile around a gene promoter, we find strong correlations between methylation and transcription changes. We also developed an extension of our method that far outperforms current approaches in identifying genes whose methylation and expression changes are correlated. With simple modifications, this tool can be used with the data we will generate in year 2 (CAGE to annotate TSSs, RNA-seq to provide expression information, and Methyl-MAPS for the high-resolution genome-wide methylation) to directly address the central hypotheses of this proposal.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 2012
- Accession Number
- ADA614053
Entities
People
- John R. Edwards
Organizations
- Washington University in St. Louis