MeDeCom
MeDeCom is a computational method within DecompPipeline for the decomposition of heterogeneous DNA methylation data. It applies matrix factorization with biologically motivated constraints and regularization to discover and quantify latent methylation components bulk DNA methylation samples. The workflow is complemented by preprocessing functionality in DecompPipeline and interactive result exploration through the FactorViz Shiny application.
Key benefits
Reference-free deconvolution of heterogeneous bulk DNA methylation profiles
Detects latent methylation components without predefined cell-type references
Uses matrix factorization with biologically motivated constraints and regularization
Includes associated preprocessing support through DecompPipeline
Supports interactive exploration of results with FactorViz
Applications
Decomposition of bulk DNA methylation data
Discovery and quantification of latent methylation components
Exploration of sample composition and methylation patterns
Visualization and interpretation of deconvolution results
Intended use
MeDeCom is intended for epigenomics researchers, bioinformaticians, computational biologists, and DNA methylation researchers working with heterogeneous methylation datasets. It is particularly suited for users who need reference-free deconvolution of bulk DNA methylomes and want to identify underlying methylation components for downstream biological interpretation.
Contact:
Website