[BioC] Limma: average over replicate spots in an MA object

[Matthew Ritchie]

Hi Mick,

If the replicate spots are printed in a systematic way (eg 
side-by-side), then lmFit() can be used to do the averaging.
For example

MA <- new("MAList", list(M=matrix(rnorm(60), 10, 6), A=matrix(rnorm(60, 
8, 2), 10, 6)))

is a random MAList object of 6 arrays with 10 genes on each.  Assuming 
the duplicate spots are printed side by side in the rows of a print-tip 
group ,

aveMdups <- lmFit(MA, design=diag(1, dim(MA)[2]), ndups=2, spacing=1)

will average over the duplicate spots, with the values stored in

aveMdups$coef

You can change the 'spacing' argument if the duplicate spots are printed 
in another way.  A similar approach can be used to average the A values:

aveAdups <- lmFit(MA$A, design=diag(1, dim(MA)[2]), ndups=2, spacing=1)

Best wishes,

Matt

michael watson (IAH-C) wrote:

>Hi
>
>The title says it all really - how do I average over replicate spots
>from an MA object in limma?
>
>Thanks
>Mick
>