Probabilistic inference of epigenetic age acceleration from cellular dynamics.

Area of Science:

• Gerontology
• Epigenetics
• Computational Biology

Background:

• Epigenetic clocks have quantitatively measured biological aging but lack mechanistic basis.
• Current epigenetic predictors are susceptible to confounding factors, complicating interpretation.
• Understanding the biological mechanisms of methylation dynamics is crucial for accurate aging assessment.

Purpose of the Study:

• To develop a mechanistic, probabilistic model for methylation transitions at the cellular level.
• To identify and quantify components of cellular dynamics that confound epigenetic aging predictors.
• To investigate the association of these components with physiological traits and aging.

Main Methods:

• Developed a probabilistic model for cellular methylation transitions.
• Identified and measured 'acceleration' and 'bias' as key components of methylation dynamics.
• Applied the model to 15,900 participants from the Generation Scotland study.
• Conducted a genome-wide association study (GWAS) of epigenetic age acceleration.

Main Results:

• The model revealed 'acceleration' and 'bias' as distinct processes confounding epigenetic predictors.
• 'Acceleration' reflects increased methylation transition speed; 'bias' reflects global methylation changes.
• Acceleration and bias showed improved associations with smoking and alcohol consumption, respectively.
• GWAS identified seven genomic loci associated with epigenetic age acceleration.

Conclusions:

• The developed probabilistic model provides a mechanistic framework for understanding epigenetic aging.
• 'Acceleration' and 'bias' are critical, distinct factors that improve the interpretation of epigenetic clocks.
• These findings advance the understanding of biological aging and its determinants, impacting geroscience research.