mmcmcBayes 0.2.0
Major updates
- Removed the Anderson–Darling test from the DMR detection
pipeline.
The decision process is now fully Bayes factor based, improving both
conceptual clarity and computational stability.
- Improved computational efficiency of the ASGN model fitting and the
multistage MCMC algorithm, resulting in faster run times across all
simulated scenarios.
- Removed the
return_mcmc option and deprecated internal
MCMC return paths in both mmcmcBayes() and
asgn_func(), simplifying the user facing interface.
New functions
- Added
summarize_dmrs() to provide a clean summary table
of detected regions, including CpG ranges, region size, decision values,
and stage information.
- Added
plot_dmr_region() for visualization of mean
M-value curve across detected DMRs. This allows users to inspect cancer
vs. normal signal patterns in a region-specific manner.
Improvements
- Cleaned and reorganized internal helper functions for clarity and
efficiency.
- Updated documentation across all exported functions, reflecting
removal of the Anderson–Darling test and removal of the
return_mcmc options.
- Enhanced the package DESCRIPTION and added examples to exported
functions.