Dissecting epigenetic heterogeneity in single-cell DNA methylomes with a unified framework

Area of Science:

• Epigenetics
• Computational Biology
• Genomics

Background:

• Single-cell DNA methylation (scDNAm) sequencing reveals epigenetic heterogeneity.
• Existing analytical methods struggle with scDNAm data's missing values and unique distributions.
• A need exists for specialized tools to analyze scDNAm data effectively.

Purpose of the Study:

• To develop a versatile analytical toolkit, scMethCraft, for systematic exploration of DNA methylation-derived epigenetic variation at single-cell resolution.
• To accurately model scDNAm data by integrating genomic features and methylation patterns.
• To facilitate downstream analyses for characterizing scDNAm heterogeneity.

Main Methods:

• scMethCraft integrates multi-perspective genomic sequence features and position information of methylated regions.
• A hybrid neural network is employed for iterative quantification of cell-to-cell associations.
• The toolkit supports cell embedding, data integration, cell type annotation, and differential methylation analysis.

Main Results:

• scMethCraft accurately reconstructs DNA methylation landscapes and models scDNAm data.
• The toolkit successfully facilitates identification and characterization of scDNAm heterogeneity.
• Integration with other workflows enabled discovery of biological insights and identification of novel oligodendrocyte-associated genes.

Conclusions:

• scMethCraft provides a robust framework for analyzing scDNAm data, overcoming limitations of existing methods.
• The toolkit enhances the discovery of epigenetic variation and regulatory mechanisms at single-cell resolution.
• scMethCraft offers an epigenetic perspective for investigating biological mechanisms in specific cellular subpopulations.