Skip to main content

Prediction of Histone methylation Patterns during the Embryonic Stem (ES) cell differentiation

Wednesday, September 16, 2015 — Poster Session I

3:30 p.m. – 5:00 p.m.
FAES Terrace


  • KD Ko
  • A Juan
  • S Wang
  • H Zare
  • V Satorelli


Histone methylation is one of well-known epigenetic mechanisms that play major roles in controlling transcription regulation. In the ES cells, transcription regulation is one of the key elements that maintains the identity cells and defines the future fate of ES cells. Identification of the core factors that govern transcriptional regulation provides valuable insights into lineage determination during ES cell differentiation. In this study, we are interested in investigating one of the histone modifications, namely H3K27 methylation. H3K27 mrthylation is catalyzed by Polycomb-group PRC2 (Polycomb Repressive Complex 2) complex to become repressive marks H3K27me2 and H3K27me3. These repressive marks are critical for proper ES cell differentiation. PRC2 in mammals contains components of EZH2 (Enhancer of zeste homolog 2). EZH2/EZH1 carries the catalytic unit for H3K27 leads to repress gene expression. However, the underlying mechanisms of why PRC2 deposit the repressive marks in specific genomic regions are not fully revealed yet. Here we manipulated the activity of PRC2 by generating a mutated form of Ezh2 in ES cells that only catalyzes H3K27me2 to H3K27me3. To investigate the key factors that decide dimethylation and trimethlyation of H3K27, we analyze mass parallel sequencing data from wide-type and mutant ES cells to collect unique features related to epigenetic modifications and transcription factor (TF) availability. Then, we will implement a pipeline to predict the dynamic changes of histone methylation in the ES cells using machine learning algorithms such as support vector machine (SVM) and deep learning.

Category: Computational Biology