[BioC] [Limma] lmFit handle missing value in phenotype

[James W. MacDonald]

Hi Chen,

On 4/4/2012 3:31 PM, yao chen wrote:
> Hi,
>
> I am using Limma to get differential expressed genes between case and
> control samples. In addition, I want to adjust the "weight" of patients by
> including it in the "design" matrix. However, one weight value of a patient
> is missing, so I have a "NA" in design matrix. When I run lmFit, I got the
> error message :Error in qr.default(x) : NA/NaN/Inf in foreign function call
> (arg 1).
>
> Can limma handle the missing value in design matrix or I have to remove
> this sample ?

It's not clear what you mean by weight here. Do you mean to say that you 
have weighed these subjects and want to control for their weight in your 
model? If I assume that this is what you mean, then you will not be able 
to include that person. If you use model.matrix() to create your design 
matrix, it will automatically drop that subject.

Best,

Jim


>
> Thanks,
>
> Chen
>
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at r-project.org
> 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
University of Washington
Environmental and Occupational Health Sciences
4225 Roosevelt Way NE, # 100
Seattle WA 98105-6099