[BioC] swaping values in a RGList object...

[Richard Pearson]

I guess this depends on exactly what you mean by "the samples were
mislabeled".

My understanding is that RG$R should always hold values for the red (Cy5)
dye, regardless of whether this is sample or reference. The targets should
say which dye is sample and which is reference. If you modify your targets
file, this will have an effect on any subsequent linear models you build,
but will make no diference to, for example, any normalisations or
subsequent MA plots. If you have simply read the Cy3 values into RG$R and
the Cy5 values into RG$G, swapping the values for the two columns will
ensure that subsequent MA plots will be as expected. Could someone with
more knowledge than myself confirm that this is correct, i.e. that the
targets file has no effect on any normalisations?

This raises an issue I've been dealing with. I have an experiment with two
sets of time series data. Each series has 17 time points. All slides have
the sample in one channel and a common (across all 34 time points)
reference in the other. For the first 17 slides, the reference is labelled
with Cy5, and in the second 17 with Cy3 (i.e. I have a dye swap). Am I
right in saying that the MA plots for the second series should be "upside
down" with respect to the first series?

If I want to make a direct comparison between the two series I should
multiply the M values of one series by -1, is that correct? Would this
still be valid after any type of within array normalisation? Would this
still be valid after any between array normalisation, or should I "swap"
the R and G values before normalising?

Many thanks

Richard


Quoting Marcus Davy <MDavy at hortresearch.co.nz>:

>
> Another way would be to modify/correct the "targets" descriptor 
> information which relates the treatments to the Cy3 and Cy5 channels 
> for each slide without having to reorder intensity information within 
> the GRList at all.
>
> Targets information is in section 5.2 of the limma users guide, 
> limmaUsersGuide().
>
> Marcus
>
>>>> Richard Pearson <Richard.Pearson at postgrad.manchester.ac.uk> 
>>>> 1/07/2005 1:19:26 p.m. >>>
> Lourdes,
>
> Try the following:
>
> R1 <- RG$R[,1]
> R2 <- RG$R[,2]
> RG$R[,1] <-R2
> RG$R[,2] <-R1
>
> Regards
>
> Richard
>
>
> Quoting Lourdes Peña Castillo <lourdes.pena at gmail.com>:
>
>> Hello Everyone,
>>
>> I am analyzing micro-array data using limma, and I need to swap
>> intensity data between arrays  (because the samples were mislabeled) .
>> Something like:
>>
>> R1 <- RG[,1]$R
>> R2 <- RG[,2]$R
>> RG[,1]$R <-R2
>> RG[,2]$R <-R1
>>
>> But I get the message:
>>
>>  Error: incorrect number of subscripts on matrix
>>
>> I have also tried replacing the values using  a for-loop but got the
>> same error.
>>
>> for (i in 1:length(RG[,1]$R)) {
>>  RG[,1]$R[i] <- R2[i]
>> }
>>
>> Do you know how I can swap the values?
>>
>> Thanks!
>>
>> Lourdes
>>
>> _______________________________________________
>> Bioconductor mailing list
>> Bioconductor at stat.math.ethz.ch
>> https://stat.ethz.ch/mailman/listinfo/bioconductor
>>
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
>
>
> ______________________________________________________
>
> The contents of this e-mail are privileged and/or confidenti...{{dropped}
}
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
>