Data ECCB 2016 main conference

PT47 – Higher order methylation features for clustering and prediction in epigenomic studies


Mississippi September 7, 2016 2:00 pm - 2:20 pm

Bookmark and Share


Proceeding talk – Theme: Data.

Abstract

DNA methylation is an intensely studied epigenetic mark, yet its functional role is incompletely understood. Attempts to quantitatively associate average methylation to expression yield poor correlations outside of the well-understood methylation-switch at CpG-islands. We propose a probabilistic model to extract higher-order features associated with the methylation profile across a defined region. These features quantitate precisely notions of shape of methylation profiles, capturing spatial correlations in methylation patterns. Using these features, we construct a powerful machine learning predictor of gene expression, significantly improving upon predictive power of average methylation. Furthermore, we use these features to cluster promoter-proximal regions, showing that five major patterns of methylation occur across cell lines, and provide evidence that methylation beyond CGIs may be related to transcriptional regulation. Results support previous reports of a functional role of spatial correlations in methylation patterns, and provide a mean to quantitate such features for downstream analyses.

Link to PDF file

Authors

Chantriolnt-Andreas Kapourani, University of Edinburgh, United Kingdom
Guido Sanguinetti, University of Edinburgh, United Kingdom