[BioC] duplicate cor?

[Jianping Jin]

Dear Bioconductor list:

Could anyone provide a good explanation on "duplicateCorrelaton"? There are 
two one-channel non-Affymetrix data sets in each of which each gene was 
spotted twice on each chip. To estimate correlation between duplicate 
values, I ran regular cor.test for each gene across all chips. The results 
showed average cor values as 0.52 and 0.81 for the two sets respectively.

The $consensus.correlation after running "duplicateCorrelation" was shown, 
however, at opposite direction: 0.57 and 0.27 (even negative in some other 
cases). I was told once that duplicateCorrelation is the residual 
correlation. But I have no idea of how to coordinate two into a clear 
picture. We know the lower is residue the better the data are correlated 
(more consistent). Does this hold on residue correlation compared to total 
correlation as well?

The on-line document on duplicateCorrelation said "consensus correlation is 
the trimmed mean of the individual correlations on the atanh-transformed 
scale". I don't know what is the atanh scale. But is it related to where I 
misunderstood the consensus correlation?

Thanks in advance!

Jianping


xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
x Jianping Jin Ph.D.            x
x Bioinformatics scientist      x
x Center for bioinformatics     x
x 3133 Bioinformatics Building  x
x CB# 7104                      x
x University of North Carolina  x
x Chapel Hill, NC 27599         x
x Tel: (919)843-6105            x
x Fax: (919)843-3103            x
x E-mail: jjin at email.unc.edu    x
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx