[BioC] contrast matrix

[James W. MacDonald]

Hi Erika,

Erika Melissari wrote:
> Dear Bioconductor list,
> 
> I am performing a microarray data analysis of a reference design
> experiment. This is a part of the target file:
> 
> Cy3 Cy5 M1775 Pool M1775 Pool M1775 Pool M1775 Pool M1775 Pool WT
> Pool WT Pool WT Pool WT Pool WT Pool A1789 Pool A1789 Pool A1789 Pool
>  A1789 Pool A1789 Pool
> 
> 
> I have four class of samples: Pool, always labeled in red, and M1775,
> A1789 and WT, always labeled in green. Then I obtain my design matrix
> by using modelMatrix ():
> 
>> design<-modelMatrix(targets,ref="Pool")
> 
>> design
> A1789 M1775 WT [1,]     0    -1  0 [2,]     0    -1  0 [3,]     0
> -1  0 [4,]     0    -1  0 [5,]     0    -1  0 [6,]     0     0 -1 
> [7,]     0     0 -1 [8,]     0     0 -1 [9,]     0     0 -1 [10,]
> 0     0 -1 [11,]    -1     0  0 [12,]    -1     0  0 [13,]    -1
> 0  0 [14,]    -1     0  0 [15,]    -1     0  0
> 
> I am interested in comparing M1775 vs WT and A1789 vs WT and to do
> this I define a contrast matrix by using:
> 
>> cont.matrix<-makeContrasts(M1775vsWT=M1775-WT,A1789vsWT=A1789-WT,levels=design)
>>  cont.matrix
> Contrasts Levels  M1775vsWT A1789vsWT A1789         0         1 M1775
> 1         0 WT           -1        -1
> 
> 
> Please, I would like to know if this is the right procedure to define
> the contrast matrix, considering that my reference sample is labeled
> in red and not in green, as in LIMMA manual examples. Please, another
> question. Why most of the time in microarray experiments do I find
> the reference sample labeled in green rather than in red? Is there
> any experimental explanation for that? Thank you very much

Yes, it is the correct procedure. Your M1775 estimates the M1775/pool
ratio, and WT estimates WT/pool, so M1775 - WT algebraically ends up
being M1775/WT.

If I am not mistaken (and it has been years since I did two color work), 
the Cy5 dye is less photo labile than the Cy3 dye, so the convention is 
usually to tag the samples you care about with the less photo labile dye.

Best,

Jim


> 
> Erika
> 
> 
> [[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