[BioC] RMA for bad chips

[James W. MacDonald]

Jai Prakash Mehta wrote:
> Hi all,
> 
> Since RMA takes all chips at a time for normalization and if the 
> datasets contain a few bad chips(chips with high background or some 
> other problem), will it affect the intensities of other good chips.

No, it won't affect the intensities of the other good chips. Remember 
that rma uses medianpolish for the model fit, so the outlier chips 
should not have any influence in the fit.

However, if you look at the residuals for the high background chips 
(using rmaPLM() in the affyPLM package and the image() function), you 
will see that the fit is not particularly good for these chips. We have 
found that the best recourse is to re-fragment the cRNA and re-hyb to a 
new chip. The vast majority of the time this results in a PM 
distribution very close to the other 'good' chips. Unfortunately you 
have to buy another chip to do so.

Best,

Jim


> 
> If so what is the alternative.
> 
> 
> Thanks and Regards
> 
> Jai Prakash Mehta
> Bioinformatics Group
> National Institute for Cellular Biotechnology
> Dublin City University
> http://www.nicb.ie
> 
> Jai Prakash Mehta
> Bioinformatics Group
> National Institute for Cellular Biotechnology
> Dublin City University
> http://www.nicb.ie
> 	[[alternative HTML version deleted]]
> 
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor


-- 
James W. MacDonald
University of Michigan
Affymetrix and cDNA Microarray Core
1500 E Medical Center Drive
Ann Arbor MI 48109
734-647-5623



**********************************************************
Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues.