[BioC] limma question

[James Wettenhall]

Hi Thomas,

> Hi
> I am working with a chip which contains "empty" spots, i.e. nothing has
> been spotted there, "landing lights", and actual spots "cDNA". I would
> like to exclude "empty" and "landing light" spots from my dataset before
> normalisation.

I think it is more common to exclude these sorts of spots after
normalization (but before fitting a linear model).

> normdata = normalizeWithinArrays(raw2)
>
> I get this
>>Error: subscript out of bounds

normalizeWithinArrays uses print-tip-group loess normalization by default.
 Because you did not specify the printer layout as an argument of
nomalizeWithinArrays(raw2), the layout is expected to be stored in:
raw2$printer.

The normalizeWithinArrays function may expect your total number of genes
to be the product of the components in the printer layout, i.e.
ngrid.r x ngrid.c x nspot.r x nspot.c
which is not true if you have excluded spots.

Hope this helps,
James