[BioC] Limma batch effect

Naomi Altman naomi at stat.psu.edu
Fri Nov 9 15:31:58 CET 2007


The strain effect is confounded with the batch effect.  As a result, 
you cannot determine whether differences are due to strain or to batch.

--Naomi

At 10:27 PM 11/8/2007, Jason Pear wrote:

>Hi Dear BioC
>I have question about design matrix considering the batch effect 
>using limma.  The experiment design is a factorial design.
>  Experiment (Affy chip)
>         Array     Strain  Treatment  Replicate  Batch
>         1        A        No         1          1
>         2        A        No         2          1
>         3        A        No         3          1
>         4        A        Yes        1          1
>         5        A        Yes        2          1
>         6        A        Yes        3          1
>         7        B        No         1          2
>         8        B        No         2          2
>         9        B        No         3          2
>         10       B        Yes        1          2
>         11       B        Yes        2          2
>         12       B        Yes        3          2
>  Here we want to konw which genes respond differently in treated 
> vs. untreated, and we also want to find out the 
> interaction.   However, the cluster analysis showed that there is a 
> clear batch effect.  So I designed the following design matrix to remove the
>batch effect:
>design
>              A_unTrt  A_Trt    B_unTrt  B_Trt    Batch
>         1    1        0        0        0        0
>         2    1        0        0        0        0
>         3    1        0        0        0        0
>         4    0        1        0        0        0
>         5    0        1        0        0        0
>         6    0        1        0        0        0
>         7    0        0        1        0        1
>         8    0        0        1        0        1
>         9    0        0        1        0        1
>         10   0        0        0        1        1
>         11   0        0        0        1        1
>         12   0        0        0        1        1
>
>contrast <- makeContrasts(
>        TrtvsUntrt=(A_Trt+B_trt)-(A_unTrt+B_unTrt),
>        interaction=(A_Trt-A_unTrt)-(B_Trt-B_unTrt),
>        levels=design)
>However, when I tried to fit the data and the following happes
>  fit <- lmFit(myData, design)
>Coefficients not estimable: Batch
>I am using limma version 2.10.5 and R version 2.5.1
>
>Any suggestions?
>
>Many thanks,
>
>Jason
>_________________________________________________________________
>[[replacing trailing spam]]
>
>         [[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

Naomi S. Altman                                814-865-3791 (voice)
Associate Professor
Dept. of Statistics                              814-863-7114 (fax)
Penn State University                         814-865-1348 (Statistics)
University Park, PA 16802-2111



More information about the Bioconductor mailing list