CHANGES IN VERSION 0.99
----------------------

NEW FEATURES

  o Initial release.

CHANGES IN VERSION 0.99-1
----------------------
  o Fixed a bug in newSeqCountSet in dealing with the input experimental designs.
  o Fixed a bug in computing local FDR. 
  o Add options to model the relationship between dispersion and mean expression.

CHANGES IN VERSION 1.2.0-1
----------------------
  o Fixed a bug in computing local FDR. 
  o Change the way to deal with genes with all 0 counts. Now the FDRs for these genes 
    are assigned as 0. 

CHANGES IN VERSION 1.6.0
----------------------
  o Implemented methods for detecting differentially methylated loci (DML). 

CHANGES IN VERSION 1.7.0
----------------------
  o Added functionalities for multiple factor experimental design. 
    Currently edgeR functions are used for glm fitting and hypothesis testing.
    DSS only provide functions for dispersion estimation.

CHANGES IN VERSION 2.0.0
----------------------
  o Implemented methods for detecting differentially methylated regions (DMR).
  o Modified the algorithm for computing variances with smoothing. 
  o Add function to detect DML for general experimental design.
  o Add function to detect DML for general experimental design.

CHANGES IN VERSION 2.5.0
----------------------
  Modify the parallel computing part in DML testing.
  We now use the 'mcapply' function in 'parallel'
  package to implement parallelization. Note that older version of DSS
  (<2.4x) used the 'bplapply' function in 'BiocParallel'
  package. However, that function somehow has significantly reduced
  performance in the new release (>1.25), so we switched to mcapply.
  A drawback is that the progress bar cannot be displayed under the
  paralelle computing setting.