2016-01-24: Ver 0.1-2
  * Fix tons of warnings and notes from R CMD check.

2014-07-24: Ver 0.1-2
  * Incorporate Cedric's codes for restarting and multiple chains.
  * Add man pages for Cedric's functions.
  * Add Cedric's codes to rename b.Mat.

2014-06-26: Ver 0.1-2
  * New work flows.

2014-06-24: Ver 0.1-1
  * Include Cedric's code.

2014-05-30: Ver 0.1-1
  * Add bias.Phi to simulation.

2014-05-19: Ver 0.1-1
  * Add bias random walk, adjust prior data structures, and modify logL.

======================= Log Before Release =====================================
2014-05-16: Ver 0.1-1
  * Add bias adjustment.

2014-05-15: Ver 0.1-0
  * Add random walk back...

2014-05-05:
  * Change random walk to M-H algorithm with restriction to hyper-parameters.

2014-04-30:
  * Add log mixture normal to Phi distribution.

2014-04-28:
  * Add random walk with restriction to hyper-parameters.

2014-04-22:
  * Add ROC + NSEf work flows.

2014-04-17:
  * Implement RW_Norm to get rid of independent chain via VGAM.

2014-04-14:
  * Add NSEf work flows.

2014-03-11:
  * Clean and reformat all work flows.

2014-02-26:
  * Add MG's methods lognormal_MG and lognormal_MG0 back, but it still not work.
  * Add trace of logL to work flows.

2014-02-21:
  * Add two component mixture normal to fit Yassour 2009 using constrOptim().
  * Add simulating expression from two component mixture fits.
  * Add above to simulation work flows.

2014-02-20:
  * Add Yassour 2009 dataset and man page.
  * Add ASL related functions and man page.

2014-02-19:
  * Remove MG's method since it did not work.
  * Add a misc code to extract parameters in work flows.

2014-02-07:
  * Add MG's proposal method for expected expression.

2014-01-16:
  * Finish conversion and debugging.
  * Add negative selection conversion to work flows.

2014-01-09:
  * Rename variables and functions (everything) to be consistent with papers.

2014-01-03:
  * Add vignettes.

2013-12-30:
  * Add with and without phi work flows.
  * Modify shell scripts.
  * Incorporate all work flows into cubfits.
  * Drop dependence of seqinr and VGAM.

2013-12-27:
  * Add man pages.

2013-12-24:
  * Test all other 24 genomes.

2013-12-14:
  * Add simulation work flow.

2013-12-13:
  * Add examples.
  * Merge back 'cubds' code to simplify dependency.

2013-11-15:
  * Add simulation code.

2013-11-14:
  * Copy input/output to "cubds".

2013-11-07:
  * Add SCUO code.

2013-10-30:
  * Add MLE and posterior mean for initial values of testing data.

2013-10-17:
  * Rename to 'cubfits'.
  * Rewrite several nasty functions.
  * Add ROC, NSEf, and ROC+NSEf models.

2013-09-xx:
  * Start a 'cub' package.
