[Bioc-sig-seq] Normalization paired samples

[Wolfgang Huber]

Dear Mayte

for some reason, the plots came as really small JPEG-images, and I am 
afraid I cannot see very much. Can you resend with better resolution?
(Perhaps use the "png" device instead of "pdf" in your R code).

Looking at the data patient-wise might, perhaps, give an indication of 
data quality issues that cannot be easily 'normalised away' and that 
might be the reason for the artefact you report.

	Best wishes
	Wolfgang


Mayte Suarez-Farinas scripsit 29/11/10 22:18:
> Dear Wolfgang
>
> I am attaching the plots with highlighted DEG tags before and after 3
> different normalization procedures, using all samples. In the second
> figure., I include the plots for the 3 patients separately before and
> after TMM normalization
> It looks like the normalization for patient 2 is bad. any idea how I can
> solve this?
>
>
> pdf('Normalization Plots.pdf',height=10,width=10)
> layout(matrix(1:4,2,2,byrow=TRUE))
> a<-plotSmear(PScounts,
> panel.first=grid(),smooth.scatter=FALSE,main='before
> normalization',de.tags=deg.tgw)
> ma.plot(a$A,a$M,plot.method='add',cex=0)
> b<-plotSmear(d.PS, panel.first=grid(),smooth.scatter=FALSE,main='after
> TMM',de.tags=deg.tgw)
> ma.plot(b$A,b$M,plot.method='add',cex=0)
> rm(b)
> b<-plotSmear(d.PS.2, panel.first=grid(),smooth.scatter=FALSE,main='after
> RLE',de.tags=deg.tgw)
> ma.plot(b$A,b$M,plot.method='add',cex=0)
> rm(b)
> b<-plotSmear(d.PS.3, panel.first=grid(),smooth.scatter=FALSE,main='after
> quantile',de.tags=deg.tgw)
> ma.plot(b$A,b$M,plot.method='add',cex=0)
> rm(b)
> dev.off()
>
> pdf('Normalization Plots TMM by pairs.pdf',height=10,width=14)
> layout(matrix(1:6,2,3,byrow=TRUE))
> a<-plotSmear(PScounts[,c(1,4)],
> panel.first=grid(),smooth.scatter=FALSE,main='Patient1 before',
> de.tags=deg.tgw, ylim=c(-10,10))
> ma.plot(a$A,a$M,plot.method='add',cex=0)
> rm(a)
> a<-plotSmear(PScounts[,c(2,5)],
> panel.first=grid(),smooth.scatter=FALSE,main='Patient2 before',
> de.tags=deg.tgw, ylim=c(-10,10))
> ma.plot(a$A,a$M,plot.method='add',cex=0)
> rm(a)
> a<-plotSmear(PScounts[,c(3,6)],
> panel.first=grid(),smooth.scatter=FALSE,main='Patient3 before',
> de.tags=deg.tgw, ylim=c(-10,10))
> ma.plot(a$A,a$M,plot.method='add',cex=0)
> rm(a)
> b<-plotSmear(d.PS[,c(1,4)],
> panel.first=grid(),smooth.scatter=FALSE,main='Patient1 after',
> de.tags=deg.tgw, ylim=c(-10,10))
> ma.plot(b$A,b$M,plot.method='add',cex=0)
> rm(b)
> b<-plotSmear(d.PS[,c(2,5)],
> panel.first=grid(),smooth.scatter=FALSE,main='Patient2 after',
> de.tags=deg.tgw, ylim=c(-10,10))
> ma.plot(b$A,b$M,plot.method='add',cex=0)
> rm(b)
> b<-plotSmear(d.PS[,c(3,6)],
> panel.first=grid(),smooth.scatter=FALSE,main='Patient3 after',
> de.tags=deg.tgw, ylim=c(-10,10))
> ma.plot(b$A,b$M,plot.method='add',cex=0)
> rm(b)
> dev.off()
>
>
> On a completely unrelated note, I had saw the video already, hilarious
> It's a pilot to see if we want to move to this technology right now, so
> it's difficult to convince my bio-collaborator to go beyond 3 right now.
> So far, some important genes that we do not find in affy arrays because
> low intensity, also came with zero counts here, even though we used
> 80bp. And it's great for me to get familiar with teh analysis issues, as
> ou can see...
>
> Mayte Suarez-Farinas
> Research Associate, The Rockefeller University
> Biostatistician, The Rockefeller University Hospital
> 1230 York Ave, Box 178,
> New York, NY, 10065
> +1(212) 327-8213
>
>
>
> On Nov 24, 2010, at 7:50 PM, bioc-sig-sequencing-request at r-project.org
> <mailto:bioc-sig-sequencing-request at r-project.org> wrote:
>
>>
>> Dear Mayte
>>
>> How do these plots look like when you make them separately for each
>> subject? (In addition, you could colour the dots according to whether
>> the differential expression analysis for the overall dataset calls them
>> 'significant').
>>
>> Also, if you compute the M values for each patient separately, how does
>> the pairs plot (scatterplot matrix) look like?
>>
>> On a completely unrelated note, I recently saw a movie about studies
>> with 3 patients:http://www.xtranormal.com/watch/6878253
>>
>> Best wishes
>> Wolfgang
>


-- 


Wolfgang Huber
EMBL
http://www.embl.de/research/units/genome_biology/huber