[BioC] limma design (paired and factorial?)

[Gordon K Smyth]

On Thu, 22 Nov 2012, maria traka (IFR) wrote:

> Thanks Gordon,

> It works brilliantly! I have a hard copy of a previous limma user guide 
> which did not contain that section. That will teach me...Sorry for 
> taking up your time!
>
> Is there a minor mistake in there? It reads block=Person when I think it 
> should read block=Patient (from the targets frame).

Yes, thanks.

Gordon

> Best wishes,
> Maria
>
> Maria Traka, PhD, MSc
> Food & Health Programme Science Manager,
> Institute of Food Research, NR4 7UA, UK
> Tel: +44 (0) 1603 255194 Fax: +44 (0) 1603 507723
> e-mail: maria.traka at ifr.ac.uk
>
> www.ifr.ac.uk www.foodandhealthnetwork.com
> e-disclaimer at http://www.ifr.ac.uk/edisclaimer/
>
> PLEASE support our moustaches for PROSTATE CANCER RESEARCH:
> http://uk.movember.com/team/683817
>
>
> -----Original Message-----
> From: Gordon K Smyth [mailto:smyth at wehi.EDU.AU]
> Sent: 22 November 2012 04:50
> To: maria traka (IFR)
> Cc: Bioconductor mailing list
> Subject: limma design (paired and factorial?)
>
> Dear Maria,
>
> If you haven't already, install the latest official release of Bioconductor.  Then look at Section 8.7 "Multi-level experiments" in the limma User's Guide, which deals with experimental designs like the one you are analysing.
>
> Best wishes
> Gordon
>
>> Date: Tue, 20 Nov 2012 12:39:33 +0000
>> From: "maria traka (IFR)" <maria.traka at ifr.ac.uk>
>> To: "bioconductor at r-project.org" <bioconductor at r-project.org>
>> Subject: [BioC] limma design (paired and factorial?)
>>
>> Dear list,
>
>> I am not sure how to create the proper design in limma for my
>> experiment which I think is a factorial and paired combined.
>>
>> I have 9 patients that are on 3 different diets (3 patients each) and
>> I have paired samples (pre and post) for each. So I have a total of 18
>> Affy arrays.
>>
>> I want to mainly determine the genes that are affected in each diet.
>> Then I also want to get the genes that are changing in the diets at
>> the 'pre' stage to get an indication of the variation in my starting
>> population. So I have made a targets file that looks like this:
>> targets
>
>>      ArrayNames Person    Diet Time
>> 1  JALI-173_post    173 Control post
>> 2   JALI-173_pre    173 Control  pre
>> 3  JALI-205_post    205   lowFV post
>> 4   JALI-205_pre    205   lowFV  pre
>> 5  JALI-223_post    223 Control post
>> 6   JALI-223_pre    223 Control  pre
>> 7  JALI-225_post    225  highFV post
>> 8   JALI-225_pre    225  highFV  pre
>> 9  JALI-235_post    235   lowFV post
>> 10  JALI-235_pre    235   lowFV  pre
>> 11 JALI-245_post    245   lowFV post
>> 12  JALI-245_pre    245   lowFV  pre
>> 13 JALI-252_post    252  highFV post
>> 14  JALI-252_pre    252  highFV  pre
>> 15 JALI-263_post    263  highFV post
>> 16  JALI-263_pre    263  highFV  pre
>> 17 JALI-276_post    276 Control post
>> 18  JALI-276_pre    276 Control  pre
>>
>>
>> then,
>>
>> person<-factor(targets$Person)
>>
>> diet<-factor(targets$Diet, levels=c("highFV","lowFV","Control"))
>>
>> time<-factor(targets$Time, levels=c("Pre", "Post"))
>>
>> So I am kind of stuck with the design and the model to use for my data
>> and also how to make contrasts and get the comparisons I want.
>>
>> Please can you give me any help?
>> Thanks in advance.
>> Maria
>>
>>
>>
>> Maria Traka, PhD, MSc
>> Food & Health Programme Science Manager, Institute of Food Research,
>> NR4 7UA, UK
>> Tel: +44 (0) 1603 255194 Fax: +44 (0) 1603 507723
>> e-mail: maria.traka at ifr.ac.uk<mailto:maria.traka at bbsrc.ac.uk>
>>
>> www.ifr.ac.uk www.foodandhealthnetwork.com
>
> ______________________________________________________________________
> The information in this email is confidential and inte...{{dropped:11}}