Entering edit mode
David
▴
860
@david-3335
Last seen 6.7 years ago
Hi,
I'm using single channel (Cy5) arrays.
i# have background corrected the arrays, removed the Cy3 components
from
the RG object. RGfinal corresponds to a subset of the genes of
interest
(controls are removed and empty spots)
#So i apply a vsn normalisation
mat.vsn <- vsnMatrix(RG.final$R)
#I try to fit a an object by taking into account duplicate spots (2
spots, spacing=1)
fit <- lmFit(mat.vsn, design, ndups=2,
correlation=corfit$consensus,weigth=RG$weights[idx,])
ebayes <- eBayes(fit)
#I compute DE genes
a <- topTable(fit2, coef=1, adjust="BH",number=100)
Now i want to retrieve the normalized values from this genes and draw
the heatmap.
The problem is that mat.vsn still contains the duplicate spots, so
extracting the values from the normalized matrix is not easy.
How can i retrieve the normalized valued across all conditions for a
particular gene without retrieveing the duplicated spots ?? The fit
object would only return the conditions stated in design but not for
all
arrays individually.
thanks
david