[BioC] limma help - choosing an approach
Naomi Altman
naomi at stat.psu.edu
Tue Sep 12 19:23:54 CEST 2006
Dear John,
You have 2 factors - time and concentration. In a factorial design,
you would have each concentration measured at each time.
I am not quite sure I understand your design. It looks like the
control might be concentration 0, and it is measured at each
time. If so, you actually do have a factorial design, and the model
statement for this design
y~time + conc + time:conc can be used to create a design matrix for
the single channel analysis (and you do not need the material from
Federer). This is the factor effects model. You do need to include
duplicatecorrelation to handle the 2 samples on the same array.
I am not sure how Genespring would handle this, but I would be quite
surprised if it is sufficiently flexible to handle your set-up
absolutely correctly.
--Naomi
At 08:40 AM 9/12/2006, john seers \(IFR\) wrote:
>Thank you very much for the reply Naomi.
>
>So, am I right in thinking that this is not a usual approach? I have
>been told that this is a standard design and that Genespring et al
>handle it routinely. Everything I look at in limma is tantalisingly
>close to what I want but not quite. I am wondering if I am making it
>more difficult than it is?
>
>Can you tell me what you mean by the "factor effects" model? i.e. is it
>in the limma userguide?
>
>Regards
>
>
>John Seers
>
>
>
>
>---
>
>John Seers
>Institute of Food Research
>Norwich Research Park
>Colney
>Norwich
>NR4 7UA
>
>
>tel +44 (0)1603 251497
>fax +44 (0)1603 507723
>e-mail john.seers at bbsrc.ac.uk
>e-disclaimer at http://www.ifr.ac.uk/edisclaimer/
>
>Web sites:
>
>www.ifr.ac.uk
>www.foodandhealthnetwork.com
>
>
>-----Original Message-----
>From: Naomi Altman [mailto:naomi at stat.psu.edu]
>Sent: 12 September 2006 13:16
>To: john seers (IFR); bioconductor at stat.math.ethz.ch
>Subject: Re: [BioC] limma help - choosing an approach
>
>
>The only text I know that covers the two-factor factorial design with
>added control is Experimental Design by W. Federer. It is out of
>print, but should be available in a university library.
>
>In the limma context, I would use separate analysis of 2 channel data
>and use the "factor effects" model that considers each treatment
>combination to be a treatment. Limma in any case uses contrasts
>rather than computing factor effects, and you probably know which
>treatments you want to compare, so the additional controls should not
>create any problems.
>
>--Naomi
>
>At 04:57 AM 9/12/2006, john seers \(IFR\) wrote:
> >
> >Hello
> >
> >I am trying to analyse some slides but I am a bit stumped. The arrays I
> >have been given vary with time and a concentration of a substance. i.e
> >two experimental dimensions (See targets data below). What is throwing
> >me is that the zero concentration is only in the control and I cannot
> >work out what model is suitable for this. Can anybody help with some
> >advice or tips on the approach to use?
> >
> >I have looked at the factorial design and the timecourse but neither
> >seem to offer a comparison/contrast against the control. Is there a way
> >to do this? Do I have to use the "Separate Analysis of Two Channel
>data"
> >in Chapter 9?
> >
> >Any advice gratefully appreciated.
> >
> >Regards
> >
> >
> >John Seers
> >
> >
> >
> >
> > My targets file looks like this:
> >
> >
> >
> >
> >SlideNumber FileName Cy3 Cy5 Time conc
> >598 598new.gpr f100t1 Control t1 c100
> >599 599new.gpr f20t4 Control t4 c20
> >600 600new.gpr f100t4 Control t1 c100
> >617 617new.gpr f20t1 Control t1 c20
> >621 621new.gpr f20t1 Control t1 c20
> >637 637new.gpr f20t4 Control t4 c20
> >638 638new.gpr f20t1 Control t1 c20
> >639 639new.gpr f100t1 Control t1 c100
> >748 748new.gpr f20t4 Control t4 c20
> >751 751new.gpr f20t4 Control t4 c20
> >833 833new.gpr f100t4 Control t4 c100
> >835 835new.gpr f100t1 Control t1 c100
> >836 836new.gpr f100t4 Control t4 c100
> >957 957new.gpr f100t1 Control t1 c100
> >958 958new.gpr f20t1 Control t1 c20
> >
> >
> >
> >
> >
> >
> >---
> >
> >John Seers
> >Institute of Food Research
> >Norwich Research Park
> >Colney
> >Norwich
> >NR4 7UA
> >
> >
> >tel +44 (0)1603 251497
> >fax +44 (0)1603 507723
> >e-mail john.seers at bbsrc.ac.uk
> >e-disclaimer at http://www.ifr.ac.uk/edisclaimer/
> ><http://www.ifr.ac.uk/edisclaimer/>
> >
> >Web sites:
> >
> >www.ifr.ac.uk <http://www.ifr.ac.uk/>
> >www.foodandhealthnetwork.com <http://www.foodandhealthnetwork.com/>
> >
> >
> > [[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
>
>_______________________________________________
>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