[BioC] Standard error in summarization ExpressionSets

[Martin Morgan]

Richard Pearson wrote:
> Hi Rodrigo
> 
> I think the se.exprs() method disappeared a release or two ago. You can 
> still get this data by using:
> 
> assayDataElement(eset, "se.exprs")

also assayData(eset)[["se.exprs"]], which might have a more familiar 
feel to it. Both return NULL when se.exprs is not present.

Martin

> 
> There are a few methods that I am aware of that give standard errors. In 
> addition to fitPLM(), I believe you can also get standard errors from 
> RMA using:
> 
> expresso(affybatch, bgcorrect.method="rma"
> , normalize.method="quantiles", pmcorrect.method="pmonly"
> , summary.method="medianpolish")
> 
> Also, MBEI (Li and Wong) will give standard errors:
> 
> expresso(affybatch
> , normalize.method="invariantset", bg.correct=FALSE
> , pmcorrect.method="pmonly", summary.method="liwong")
> )
> 
> My personal recommendation for getting standard errors would be to use 
> mmgmos() from the package puma (though I might be biased :-)
> 
> You could also look at the function bgx() in the package of the same name
> 
> HTH
> 
> Richard.
> 
> 
> James W. MacDonald wrote:
>> Hi Rodrigo,
>>
>> Rodrigo Santamaría Vicente wrote:
>>> Hi there!,
>>> After performing normalization-summarization with any of the 
>>> available functions (rma, mas5, plier, etc.), I get an ExpressionSet 
>>> class. It inherits from eSet class, but it seems to loss the 
>>> se.exprs() method to get the standard errors for the summarized 
>>> expression levels.
>>> My question is: how could I get the standard errors from the 
>>> ExpressionSet returned by rma, mas5 or any other method? if these 
>>> methods do not return standard errors, why? are them unimportant? is 
>>> there any method that computes them?
>>
>> None of the methods you mention return standard errors. If you want 
>> standard errors, you might use the affyPLM package. For instance the 
>> fitPLM() function will return values very similar to rma(), but with 
>> standard errors.
>>
>> Best,
>>
>> Jim
>>
>>
>>>
>>> Thanks in advance!
>>> Rodrigo
>>>
>>> _______________________________________________
>>> Bioconductor mailing list
>>> Bioconductor at stat.math.ethz.ch
>>> https://stat.ethz.ch/mailman/listinfo/bioconductor
>>> Search the archives: 
>>> http://news.gmane.org/gmane.science.biology.informatics.conductor
>>
> 


-- 
Martin Morgan
Computational Biology / Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N.
PO Box 19024 Seattle, WA 98109

Location: Arnold Building M2 B169
Phone: (206) 667-2793