DMCFB

DMCFB identifies differentially methylated cytosines using a Bayesian functional regression model in bisulfite sequencing data.

Key Features:

• Functional Regression Model: Captures position-specific, group-specific, and covariates-specific methylation patterns.
• Spatial Correlation Handling: Models spatial correlation patterns and imputes missing values using spatial correlation between positions and samples.
• Bayesian Estimation: Adopts a Bayesian approach for estimation and inference.
• Model Flexibility: Remains robust and flexible with respect to true underlying models and inclusion of any covariates.

Scientific Applications:

• Differential Methylation Analysis: Identifies differentially methylated cytosines in bisulfite sequencing data.

Methodology:

DMCFB uses a Bayesian functional regression model that captures position-specific, group-specific and covariates-specific methylation patterns, spatial correlation patterns, and unknown underlying models of methylation data; missing values are imputed using spatial correlation between positions and samples.