[BioC] duplicateCorrelation problem with modelMatrix

[Sergio Barberan]

Hi,

I'm trying to use duplicateCorrelation for duplicate spots with no
success.
My array design includes three replicas of each spot, 120 spots apart so
I use:

cor <- duplicateCorrelation(MA,design,ndups=3,spacing=120)
correlation <- cor$consensus.correlation
fit <- lmFit(MA,design,ndups=3,spacing=120,correlation=correlation)

That works when I analyze a simple experiment with a design of (1,-1).
But when I have the following design:

	ad em l1 l2 l3 l4
V3_slide11  0  -1  0  0  0  0
V3_slide14  0   0 -1  0  0  0
V3_slide15  0   0  0 -1  0  0
V3_slide16  0   0  0  0 -1  0
V3_slide17  0   0  0  0  0 -1
V3_slide18 -1   0  0  0  0  0


I can not get the correlation so I have to use the default 0.75
correlation for lmFit in order to work.

Thanks in advance,
Sergio