[BioC] LIMMA: duplicate correlation

[Gordon Smyth]

>Date: Thu, 22 Sep 2005 07:38:19 +0200
>From: "Hoen, P.A.C. 't \(HKG\)" <P.A.C._t_Hoen at lumc.nl>
>Subject: [BioC] LIMMA: duplicate correlation
>To: <bioconductor at stat.math.ethz.ch>
>
>Dear all,
>I am analysing results from a spotted microarray with 9,600 features, 
>spotted in duplicate.
>When I calculate the duplicate correlation, as follows:
>
>         cor <- duplicateCorrelation(datanorm_ordered, design, ndups=2)
>
>I get the following error message:
>#5: Too much damping - convergence tolerance not achievable in: 
>glmgam.fit(dx, dy, start = start, tol = tol, maxit = maxit, trace = trace)

This is not an error message, this is a WARNING. As I have said many times 
on this list, getting a small number of warning messages from 
duplicateCorrelation() is not a problem. Ignore it.

Gordon

>The consensus correlation coefficient is low (0.100) and also the 
>distribution of correlation coefficients is between -0.1 and 0.5.
>
>Still, I seem to get reasonable results from the model fit on the 8 arrays.
>Does anyone know what the problem is here?
>
>Thanks a lot,
>Peter-Bram