[BioC] limma: duplicates handling

[Jenny Drnevich]

Hi Andres,

Did you read through the help file for 
duplicateCorrelation? You can't do both block and ndups:

"At this time it is not possible to estimate 
correlations between duplicate spots and between 
technical replicates simultaneously. If block is 
not null, then the function will set ndups=1."

If you have a blocking variable and technical 
replicates, you cannot use limma to adjust for 
both. What to do in this situation has been 
discussed previously on this list many times; see:
https://stat.ethz.ch/pipermail/bioconductor/2008-November/025116.html

HTH,
Jenny


At 08:40 AM 4/22/2009, Andres Pinzon wrote:
>Hi Jenny,
>Thanks for the advice. This is the part of the  code I think is not working:
>===
>+
>+>biolrep <- c(1,1,2,2,3,3)
>+>corfit <- duplicateCorrelation(MA,ndups=2,block=biolrep, spacing=8)
>+>fit <- lmFit (MA, block=biolrep, cor=corfit$consensus)
>+>fit <- eBayes(fit)
>+
>===
>
>Shoudl I  use something like?:
>
>fit <- lmFit (MA, block=biolrep, ndups=2, spacing=8, cor=corfit$consensus)
>
>
>
>best,
>
>
>On Wed, Apr 22, 2009 at 9:21 AM, Jenny Drnevich <drnevich at illinois.edu> wrote:
> > Hi Andres and Jim,
> >
> > Actually, there is a way that limma accounts for the duplicates and only
> > reports one value per clone in topTable. Did you follow the Within-Array
> > Replicate Spot example 11.6 in the limmaUsersGuide()? After you use
> > duplicateCorrelation() to calculate the correlation between spots, you have
> > to modify the call to lmFit:
> >
> > fit <- lmFit(MA2, design, ndups=2, correlation=corfit$consensus)
> >
> > Andres - in the future, it would help if you posted your code that wasn't
> > working along with the output of sessionInfo() so we could see what exactly
> > you did and what you need to do.
> >
> > Cheers,
> > Jenny
> >
> > At 07:59 AM 4/22/2009, James W. MacDonald wrote:
> >>
> >> Hi Andres,
> >>
> >> Andres Pinzon wrote:
> >>>
> >>> Hi James!
> >>> On Wed, Apr 22, 2009 at 8:40 AM, James W. MacDonald
> >>> <jmacdon at med.umich.edu> wrote:
> >>>>
> >>>> Hi Andres,
> >>>>
> >>>> Andres Pinzon wrote:
> >>>>>
> >>>>> Hi Everyone,
> >>>>>
> >>>>> If I have duplicates in each slide of my experiment, how do I tell
> >>>>> limma
> >>>>> to handle this?
> >>>>> I am using duplicateCorrelation() function, but after the whole process
> >>>>> the topTable() reports not half of the spots but all of them,
> >>>>> For instance, if there are overall 15000 spots in my experiment, and
> >>>>> half of them are
> >>>>> duplicates, Shouldn't I end up just with 7500 genes?
> >>>>
> >>>> Nope. When you use duplicateCorrelation() you are telling limma to fit a
> >>>> mixed model that accounts for correlation between duplicate spots. But
> >>>> you
> >>>> are not telling limma to take averages and just report one value.
> >>>
> >>> Ok, thanks for answering. Any idea on how to tell limma to take
> >>> average and report just one value?
> >>> Sorry I really can not figure it out.
> >>
> >> I don't think there are facilities within limma to do this. If this is
> >> what you want to do, then you should just 
> compute the averages and then fit
> >> the model using the averages.
> >>
> >>
> >>> Best,
> >>
> >> --
> >> 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
> >>
> >> _______________________________________________
> >> 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
> >
> > Jenny Drnevich, Ph.D.
> >
> > Functional Genomics Bioinformatics Specialist
> > W.M. Keck Center for Comparative and Functional Genomics
> > Roy J. Carver Biotechnology Center
> > University of Illinois, Urbana-Champaign
> >
> > 330 ERML
> > 1201 W. Gregory Dr.
> > Urbana, IL 61801
> > USA
> >
> > ph: 217-244-7355
> > fax: 217-265-5066
> > e-mail: drnevich at illinois.edu
> >
>
>
>
>--
>Andrés Pinzón
>http://bioinf.ibun.unal.edu.co/~apinzon/
>Bioinformatics Center, Colombia EMBnet node
>http://bioinf.ibun.unal.edu.co
>Tel +57 3165000 ext 16961 Fax +571 3165415
>Micology and Phytopathology Laboratory - Los Andes University.
>http://bioinf.uniandes.edu.co
>Tel +571 3394949 ext. 2768

Jenny Drnevich, Ph.D.

Functional Genomics Bioinformatics Specialist
W.M. Keck Center for Comparative and Functional Genomics
Roy J. Carver Biotechnology Center
University of Illinois, Urbana-Champaign

330 ERML
1201 W. Gregory Dr.
Urbana, IL 61801
USA

ph: 217-244-7355
fax: 217-265-5066
e-mail: drnevich at illinois.edu