how to run SAM analysis

0

Entering edit mode

weinong han ▴ 270

@weinong-han-1250

Last seen 9.7 years ago

An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20050831/ 1b4c3f6b/attachment.pl

• 854 views

ADD COMMENT • link updated 18.7 years ago by Adaikalavan Ramasamy ★ 1.8k • written 18.7 years ago by weinong han ▴ 270

0

Entering edit mode

Adaikalavan Ramasamy ★ 1.8k

@adaikalavan-ramasamy-675

Last seen 9.7 years ago

Your design is seriously unbalanced and under-powered. I would suggest you increase the sample size in the normal group. This holds no matter what test statistics you plan on using. LIMMA would probably be better for your purposes. I _think_ it has been shown empirically to hold for small sample sizes. Regards, Adai On Wed, 2005-08-31 at 19:22 -0700, weinong han wrote: > Hi, List, > > I have 17 affymetrix .cel files and ran RMA analysis on them. Now I want to run SAM analysis(unpaired, 3 normal tissues and 14 diseased tissues) on the RMA data with R2.1.1. > I have installed the samr package. > > Questions: How to run the SAM analysis after the RMA analysis? The better, the more detailed explanations. > > Any advice and suggestions will be much appreciated. > > > > > Best Regards > > Han Weinong > hanweinong at yahoo.com > > __________________________________________________ > > > > [[alternative HTML version deleted]] > > _______________________________________________ > Bioconductor mailing list > Bioconductor at stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor >

ADD COMMENT • link 18.7 years ago Adaikalavan Ramasamy ★ 1.8k

0

Entering edit mode

hi! if you want to run SAM anaylsis anyway, try the following (not sure if it works that way, but worth a try) for example (RMA or gcRMA as you want, for gcRMA load the library gcRMA) data.norm <- rma(data, normalize = TRUE, background = TRUE) now construct the vector cl as needed by the SAM function: cl <- c(0,0,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1) or cl <- rep(0:1, c(3,14)) with 0 being the normal tissues and 1 the diseased ones. (use pData(data) to see the correct order of your data) and then: data.sam <- sam(data.norm, cl) don?t forget to filter the genes with the genefilter function ( genefilter(data,flist) ) in the genefilter library! please everyone send me positive or negative answers, would appreciate if my instructions were correct or not. thanks! cheers, gregor

<font style="FONT-SIZE:11px;FONT-FAMILY:tahoma,sans-serif">
From: Adaikalavan Ramasamy <ramasamy at cancer.org.uk>
Reply-To: ramasamy at cancer.org.uk
To: weinong han <hanweinong at yahoo.com>
CC: bioconductor at stat.math.ethz.ch
Subject: Re: [BioC] how to run SAM analysis
Date: Thu, 01 Sep 2005 22:48:51 +0100
>Your design is seriously unbalanced and under- powered. I would suggest
>you increase the sample size in the normal group. This holds no matter
>what test statistics you plan on using.
>
>LIMMA would probably be better for your purposes. I _think_ it has been
>shown empirically to hold for small sample sizes.
>
>Regards, Adai
>
>
>
>On Wed, 2005-08-31 at 19:22 -0700, weinong han wrote:
> > Hi, List,
> >
> > I have 17 affymetrix .cel files and ran RMA analysis on them. Now I want to run SAM analysis(unpaired, 3 normal tissues and 14 diseased tissues) on the RMA data with R2.1.1.
> > I have installed the samr package.
> >
> > Questions: How to run the SAM analysis after the RMA analysis? The better, the more detailed explanations.
> >
> > Any advice and suggestions will be much appreciated.
> >
> >
> >
> >
> > Best Regards
> >
> > Han Weinong
> > hanweinong at yahoo.com
> >
> > __________________________________________________
> >
> >
> >
> > [[alternative HTML version deleted]]
> >
> > _______________________________________________
> > Bioconductor mailing list
> > Bioconductor at stat.math.ethz.ch
> > https://stat.ethz.ch/mailman/listinfo/bioconductor
> >
>
>_______________________________________________
>Bioconductor mailing list
>Bioconductor at stat.math.ethz.ch
>https://stat.ethz .ch/mailman/listinfo/bioconductor
</font>

ADD REPLY • link 18.7 years ago Gregor Siegler ▴ 50

Login before adding your answer.