## IMPORTANT 
- move IRanges and oligoClasses back to Suggests? BC moved them to
Depends just to get rid of annoying warnings while transitioning.

#####################################
### FOR CRLMM
#####################################
- Add RS ids to annotation packages
- Allele plots
- M v S plots

R. Scharpf (after next release)

- crlmm should return a SnpSet object with elements A, B, calls, and
  confs in the assayData when save.it=TRUE.  These objects can be very
  big, so...  

- crlmm function should have an option to return nothing and write to
  file a SnpSet object for each chromosome.  This would help the copy
  number part that runs by chromosome

B Carvalho

- crlmm should return results in CHP files

- crlmm2() with save.it=TRUE must save the original preprocessed intensities

  ***********************************************************
  *                                                         *       
  *                FOR CNRMA                                *
  *                                                         *
  ***********************************************************

  o Bias adjustment for X chromosome

  o Adjust nu for altered copy number (crosshyb)

  o Should store parameters in something besides the featureData slot

    - Perhaps add a matrix slot for parameters and a slot for batch.

    - Need accessors

    - how to store parameters for chromosome X (should be nu_A_male,
      nu_A_female, etc.)?

  o Provide a SNP- and sample-specific estimate of the variance for
    computing emission probabilities.  (Currently, only SNP-specific)

    (DONE -- version 1.3.10

  o Allow a copy-neutral ROH state by not having an equal prior on the
    ellipses.  For instance,

    normal state:  1/4 AA, 1/2 AB, 1/4 BB
    
    copy-neutral ROH:  (1-epsilon)/2 AA,  epsilon AB,  (1-epsilon)/2
    BB

  o Define a class that contains settings for genotyping/copynumber estimation