Testing cell-type-specific mediation effects in genome-wide epigenetic studies.

Area of Science:

• Genetics and Epigenetics
• Computational Biology
• Environmental Health

Background:

• Epigenome-wide mediation analysis identifies DNA methylation CpG sites linking exposures to health outcomes.
• Bulk DNA methylation data from blood can introduce confounding bias and reduce power due to cell heterogeneity.
• Current methods lack the ability to perform cell-type-specific mediation analysis.

Purpose of the Study:

• To develop a novel method for identifying cell-type-specific mediation effects using only bulk-level DNA methylation data.
• To address the limitations of using heterogeneous cell populations in mediation analysis.
• To provide a tool for more precise identification of CpG sites mediating exposure-outcome relationships.

Main Methods:

• Proposed a novel method, Mediation In a Cell-type-Specific fashion (MICS).
• MICS assesses exposure-mediator and mediator-outcome associations within each cell type.
• Employs the MultiMed procedure for combining cell-type-specific associations and controlling for multiple testing.

Main Results:

• Simulation studies confirmed MICS achieves correct False Discovery Rate (FDR) control.
• Applied MICS to the Normative Aging Study dataset.
• Identified nine DNA methylation CpG sites in lymphocytes mediating the effect of smoking on lung function.

Conclusions:

• MICS enables cell-type-specific mediation analysis from bulk DNA methylation data.
• The method enhances the precision of identifying mediating CpG sites in epigenome-wide studies.
• MICS offers a valuable approach for understanding cell-specific molecular mechanisms in environmental health research.