[BioC] Ringo package: biological replicate correlation within cher (chip-enriched region)

[Diego Villar]

Dear BioC community,

  I am using the Ringo package for analysis of ChIP-chip experiments on
the  Agilent platform. While the package works wonderfully, I have some
doubts regarding ordering of ChIP-enriched regions (chers) that I hope
someone can clarify. Specifically, I have four biological replicates and
I find that, while cher finding on the smoothed reporter signals forms
nice "peaks" in the found chers, biological replicate correlation of
probe intensities within the cher is not considered for ordering, and
therefore many show bad reproducibility when looking at individual
experiments. On the other hand, most combinations of parameters miss
positive control peaks that show good reproducibility among biological
replicates.

  As I am new to Bioconductor, I wonder whether Ringo functions or other
R packages can be used to evaluate biological replicate consistency in
the signals within a cher. I have skimmed through the vignettes of ACME
and CMART packages, but to my understanding p values are to be
calculated for each sample independently.

  Thanks everyone in advance for your time and help,

              Diego Villar

           Diego Villar
           Laboratorio 2.5.2
           Teléfono 915854442
           Instituto de Investigaciones Biomédicas     "Alberto Sols"
           CArturo Duperier, 4 28029 Madrid