Modeling methylation dynamics with simultaneous changes in CpG islands.

Area of Science:

• Genomics
• Computational Biology
• Epigenetics

Background:

• Vertebrate genomes feature CpG islands where methylation levels are regulated.
• Gene regulation can cause simultaneous methylation changes across multiple CpG sites within an island.
• Accurate quantification of methylation change requires accounting for these coordinated, island-wide events.

Purpose of the Study:

• To propose a probabilistic model (IWE-SSE) for methylation dynamics that considers simultaneous changes within CpG islands.
• To develop a Markov-chain Monte-Carlo (MCMC) method for fitting this model to cell type phylogeny data.
• To evaluate the impact of accounting for CpG island-wide methylation changes on phylogenetic inference.

Main Methods:

• Developed the IWE-SSE probabilistic model for methylation dynamics.
• Implemented an MCMC method for model fitting to methylation data.
• Applied the model and method to data from murine hematopoietic cells and human cell lines.
• Conducted simulation studies to validate findings.

Main Results:

• The IWE-SSE model effectively captures simultaneous methylation changes within CpG islands.
• The MCMC method accurately fits the model to empirical data.
• Accounting for CpG island-wide changes significantly impacts inferred branch lengths in phylogenies.
• The model provides a substantially improved fit for methylation data from murine and human cell studies.

Conclusions:

• The IWE-SSE model and MCMC inference method enable quantification of methylation changes at both single CpG sites and entire CpG islands.
• Incorporating island-wide methylation changes improves the accuracy of branch length estimation in phylogenetic analyses.
• This approach offers a more precise understanding of epigenetic dynamics during cell differentiation and evolution.