v2.1.1

RAMEN v2.1.1

This patch fixes some minor bugs in RAMEN including:

• Improved argument check - empty objects now throw errors, and all functions
  check for matched IDs across all objects.
• Fixed bug in nullDistGE() where the seed was fixed to 1 during the
  selectVariables() step instead of using the seed argument provided.
• Fixed a bug in selectVariables() where the function would throw an error
  when, in the absence of covariates, a VML had only 1 SNP. In that case, the
  variable selection for G is now not conducted and the same SNP is returned.
• Fixed a bug in lmGE() where the function would throw an error when no
  covariates were fed into the function.
• summarizeVML() now explicitly errors if the VML object has duplicate
  VML_index values, instead of silently merging them.

In addition to that, we added a section on the vignette explaining how to set
up a parallel backend, and expanding on the data assumptions/expectations
from RAMEN.

The IlluminaHumanMethylation450kanno.ilmn12.hg19 and
IlluminaHumanMethylationEPICv2anno.20a1.hg38 annotation packages have been
moved Suggests to lighten the default installation, since most
users only work with one microarray platform at a time.

This version also includes several internal performance improvements (faster
argument validation, VMR correlation computation, probe variability scoring,
and VML summarization), with no change in output.

This patch presents no changes on the user's end, other than the dependency
change and the summarizeVML() duplicate-index check noted above.