MethylMix

MethylMix identifies disease-specific hypermethylated and hypomethylated genes whose methylation changes predict transcriptional alterations by modeling genome-wide DNA methylation with a beta mixture model.

Key Features:

• Beta mixture model: Employs a beta mixture model to discern distinct DNA methylation states and differentiate normal from aberrant methylation patterns.
• Differential Methylation Value (DM-value): Calculates DM-values that quantify the difference in methylation state between diseased and normal conditions for each gene.
• Integration with gene expression data: Correlates matched gene expression data with methylation states to identify transcriptionally predictive methylation alterations.
• Identification of hyper- and hypomethylated genes: Detects genes exhibiting disease-specific hypermethylation or hypomethylation that are functionally relevant to transcriptional regulation.

Scientific Applications:

• Cancer research: Used to identify methylation changes implicated in carcinogenesis and tumorigenesis across different cancers.
• Biomarker discovery: Facilitates identification of differentially methylated genes that may serve as diagnostic biomarkers or molecular targets linked to altered gene expression.

Methodology:

Inputs genome-wide DNA methylation data (typically from high-throughput assays); applies a beta mixture model to classify methylation states and compare them to a baseline normal state; computes DM-values for each gene; and integrates matched gene expression data to identify methylation changes that correlate with expression alterations.