[BioC] Limma: how to extract mean of groups?

[James W. MacDonald]

Hi Guido,

Hooiveld, Guido wrote:
> Hi James,
> 
> Thanks. I indeed had seen that before, but I had the impression the
> 'coefficients' actually are the log2 of the fold change...?!?
> Anywayz, I'll have a closer look at it to make sure I understand it
> correctly.

It depends on how you set up the design matrix. If you have an 
intercept, then the coefficients will be the difference between a 
baseline level and whatever level you are looking at. However, if you 
don't have an intercept, then the coefficients represent the group means.

Best,

Jim


> 
> Best,
> Guido
> 
> 
>  
> 
>> -----Original Message-----
>> From: James W. MacDonald [mailto:jmacdon at med.umich.edu] 
>> Sent: 06 July 2009 18:50
>> To: Hooiveld, Guido
>> Cc: bioconductor at stat.math.ethz.ch
>> Subject: Re: [BioC] Limma: how to extract mean of groups?
>>
>> Hi Guido,
>>
>> Hooiveld, Guido wrote:
>>> Dear list,
>>>  
>>> Is it possible to extract the group means + SD from Limma's output?
>>>  
>>> I do know that the "Amean" can be extracted (fit2$Amean), 
>> but this is 
>>> the mean across all arrays/groups, and i am also interested in the 
>>> means of the experimental groups.
>> Have you read the relevant help page (hint ?MArrayLM-class)?
>>
>> I get
>>
>> Slots/Components:
>>
>>       'MArrayLM' objects do not contain any slots (apart from '.Data')
>>       but they should contain the following list components:
>>
>>       'coefficients': 'matrix' containing fitted coefficients or
>>            contrasts
>>
>>       'stdev.unscaled': 'matrix' containing unscaled standard 
>> deviations
>>            of the coefficients or contrasts
>>
>> Which appears to be what you want.
>>
>> Best,
>>
>> Jim
>>
>>
>>> Therefore, is it somehow possible to extact the mean (+ SD) of the 
>>> groups that are defined by the contrast matrix? Thus in my 
>> particular 
>>> case the mean + SD of the WT.Con, WT.WY. KO.Con and KO.WY groups.
>>>  
>>> Thanks,
>>> Guido
>>>  
>>>  
>>> library(affy)
>>> library(limma)
>>>  
>>> targets <- readTargets("targets.txt")
>>> data <- ReadAffy(filenames=targets$FileName)
>>> x.norm <- rma(data)
>>>
>>> TS <- paste(targets$Strain, targets$Treatment, sep=".") TS <- 
>>> factor(TS, levels=c("WT.Con","WT.WY","KO.Con","KO.WY"))
>>> design <- model.matrix(~0+TS)
>>> colnames(design) <- levels(TS)
>>> fit <- lmFit(eset, design)
>>> cont.matrix <- makeContrasts(WTwyvWTc=WT.WY-WT.Con,
>>> KOwyvKOc=KO.WY-KO.Con, levels=design)
>>> fit2 <- contrasts.fit(fit, cont.matrix)
>>> fit2 <- eBayes(fit2)
>>>
>>> ------------------------------------------------
>>> Guido Hooiveld, PhD
>>> Nutrition, Metabolism & Genomics Group Division of Human Nutrition 
>>> Wageningen University Biotechnion, Bomenweg 2
>>> NL-6703 HD Wageningen
>>> the Netherlands
>>> tel: (+)31 317 485788
>>> fax: (+)31 317 483342 
>>> internet:   http://nutrigene.4t.com <http://nutrigene.4t.com/>  
>>> email:      guido.hooiveld at wur.nl 
>>>
>>>
>>>
>>> 	[[alternative HTML version deleted]]
>>>
>>> _______________________________________________
>>> 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
>> --
>> James W. MacDonald, M.S.
>> Biostatistician
>> Douglas Lab
>> University of Michigan
>> Department of Human Genetics
>> 5912 Buhl
>> 1241 E. Catherine St.
>> Ann Arbor MI 48109-5618
>> 734-615-7826
>>
>>
> 

-- 
James W. MacDonald, M.S.
Biostatistician
Douglas Lab
University of Michigan
Department of Human Genetics
5912 Buhl
1241 E. Catherine St.
Ann Arbor MI 48109-5618
734-615-7826