[BioC] Quality checks / flagging

Yves Bastide Yves.Bastide at irisa.fr
Thu Sep 18 11:58:36 MEST 2003

Gordon Smyth wrote:
> At 05:52 PM 18/09/2003, Yves Bastide wrote:
>>>> An even more elementary question (disclaimer: I am not a 
>>>> statistician): how to use these weights to make limma ignore some 
>>>> lines?  I naively tried with weights of 0 and 1, but it didn't work 
>>>> as expected.
>>> Weights of 0 and 1 do have the effect of ignoring the log-ratios for 
>>> spots with zero weight. I am not sure what you mean by "ignore some 
>>> lines". Please explain further what you did and what you expected to 
>>> happen which didn't.
>> Oh.  I zero-weighted flagged spots and spots for unwanted genes.  This 
>> resulted in about 600 genes to consider on each array, out of 4096.
> In my honest opinion, throwing out almost all your data from a 
> microarray experiment is crazy unless you have decided that there is 
> something seriously wrong with the array. If you really want to do this, 
> then please use normalizeWithinArrays with method="loess" rather than 
> the default. You've thrown out almost all your data so there really 
> isn't enough left to expect to be able to do normalization by print-tip 
> groups.

Yep, that's the conclusion I arrived to.  My basic flawed methodology 
was to find differentially expressed genes in one group, then re-read 
the arrays for another group, lather, rince, repeat.  Gee.

> Gordon



PS: one question on limma.R (1.2.1, 1.2.2)
toptable, line 511: does the else apply to the right if?
                 if(sort.by=="A") stop("Cannot sort by A-values as these 
have not been given")
   (and yes, I'm a Python user)

More information about the Bioconductor mailing list