getting normexp (and log2 transformed) values - not further-processed M values
4
0
Entering edit mode
Last seen 7.2 years ago
Dear Listserv, I have an unusual microarray data set -a RIP-ChIP data set- and would like to use the "normexp" function (as found in limma) to background- subtract my data (and I'll take it's log2 data transformation, too) - but I don't want the automatic M and A value output that you can get from limmaGUI. [This is solely because my data are not the least bit normal and I cannot use limma to analyze them further.] Is there a way for me to obtain the normexp (and transformed) data from within R? Is is an "object" that I can "print" or "write" before it gets converted to M values? Thank you very much for your help, Sincerely, Alana
0
Entering edit mode
@sean-davis-490
Last seen 7 weeks ago
United States
On Sun, Oct 4, 2009 at 9:20 PM, <adoty at="" umd.edu=""> wrote: > > Dear Listserv, > > I have an unusual microarray data set -a RIP-ChIP data set- and would like to use the "normexp" function (as found in limma) to background-subtract my data (and I'll take it's log2 data transformation, too) - but I don't want the automatic M and A value output that you can get from limmaGUI. ?[This is solely because my data are not the least bit normal and I cannot use limma to analyze them further.] > > Is there a way for me to obtain the normexp (and transformed) data from within R? ?Is is an "object" that I can "print" or "write" before it gets converted to M values? > Hi, Alana. Have a look at the help for the backgroundCorrect() function in the limma package. Sean
0
Entering edit mode
Last seen 7.2 years ago
Dear Scott et al. Thank you very much for the information. After reading both the limma.pdf and the usersguide.pdf (and playing around with my Windows R console), I now know that by typing "fix(RG)" in limmaGUI's "Evaluate R Code" tab, I can get the RG object to display itself in some form... I do, though, have 3 remaining (and hopefully final) questions on this topic: 1) what's the difference between RG and RGList? I see them both used in the help/example files and I can't tell which one has the output I'd like from "backgroundCorrect" (-again, I'd like to obtain the normexp backgrounded-corrected values, using an offset, for both the R and G channels from each of my 5 arrays) 2) will the values in whichever of these "objects" I need [RG or RGList] be log2 transformed already? I'm not sure if this is automaticaly done or not... 3) when I look at the corrected data with "fix(RG)", it's in matrix form, right? How can I get the data into, say, a form readable by Excel? I think I can get some useful output with "show(RG)," too - but in this form it would require a lot of copying/pasting to get into Excel (and by default, not all the lines are shown). Thank you VERY much for any help someone may be able to provide. I am trying to learn what I can on my own! - but it's going a bit more slowly than I'd like some answers by... Sincerely, Alana Doty University of Maryland -------------- next part -------------- An embedded message was scrubbed... From: Sean Davis <seandavi@gmail.com> Subject: Re: [BioC] getting normexp (and log2 transformed) values - not further-processed M values Date: Mon, 5 Oct 2009 07:07:14 -0400 Size: 3187 URL: <https: stat.ethz.ch="" pipermail="" bioconductor="" attachments="" 20091005="" 26329477="" attachment.eml="">
0
Entering edit mode
@gordon-smyth
Last seen 2 hours ago
WEHI, Melbourne, Australia
Dear Alana, Since you want to do your own customized analysis of the data, which is different from what limmaGUI provides, you will find it easier to use the command-line limma package rather than limmaGUI. All objects created will be quite transparent. If you want to output values from an RGList object to a text file, you can type write.table(file="yourfilenamered.txt",RG$R) and write.table(file="yourfilenamegreen.txt",RG$G) However I find it hard to believe this is a good idea. Better to find out how to do what you want in R. I suspect you're making things harder for yourself than it needs to be. Best wishes Gordon > Date: Mon, 5 Oct 2009 19:36:38 -0400 (EDT) > From: <adoty at="" umd.edu=""> > Subject: [BioC] getting normexp (and log2 transformed) values - not > further-processed M values > To: bioconductor at stat.math.ethz.ch > Message-ID: <20091005193638.ATB49372 at po5.mail.umd.edu> > Content-Type: text/plain; charset="us-ascii" > > > Dear Scott et al. > > Thank you very much for the information. After reading both the > limma.pdf and the usersguide.pdf (and playing around with my Windows R > console), I now know that by typing "fix(RG)" in limmaGUI's "Evaluate R > Code" tab, I can get the RG object to display itself in some form... > > > I do, though, have 3 remaining (and hopefully final) questions on this topic: > > 1) what's the difference between RG and RGList? I see them both used in > the help/example files and I can't tell which one has the output I'd > like from "backgroundCorrect" (-again, I'd like to obtain the normexp > backgrounded-corrected values, using an offset, for both the R and G > channels from each of my 5 arrays) > > 2) will the values in whichever of these "objects" I need [RG or RGList] > be log2 transformed already? I'm not sure if this is automaticaly done > or not... > > 3) when I look at the corrected data with "fix(RG)", it's in matrix > form, right? How can I get the data into, say, a form readable by > Excel? I think I can get some useful output with "show(RG)," too - but > in this form it would require a lot of copying/pasting to get into Excel > (and by default, not all the lines are shown). > > > Thank you VERY much for any help someone may be able to provide. I am > trying to learn what I can on my own! - but it's going a bit more slowly > than I'd like some answers by... > > > Sincerely, > Alana Doty > University of Maryland > > -------------- next part -------------- > An embedded message was scrubbed... > From: Sean Davis <seandavi at="" gmail.com=""> > Subject: Re: [BioC] getting normexp (and log2 transformed) values - not > further-processed M values > Date: Mon, 5 Oct 2009 07:07:14 -0400 > Size: 3187 > URL: <https: stat.ethz.ch="" pipermail="" bioconductor="" attachments="" 200910="" 05="" 26329477="" attachment-0001.eml=""> > > ------------------------------
0
Entering edit mode
Last seen 7.2 years ago
Dear Dr. Smyth, Thanks so much for your quick and helpful reply. Yes, I think you are right to say "just use R!" I never had before I began to analyze my RIP-CHiP data set (hence the non-standard use of limmaGUI) - but I have indeed thoroughly read the limma user's manual (and I also tried general R help documentation and the limma "user manual" to investigate R more...) Anyway, your help allowed me to solve a similar "output" problem in another R package (eCADS). I also see that by using the "exprs.MA(MA)" statement, I can recover normalized, log2 transformed individual values for both the red and green channels for all my arrays. Thank you very much again for your help. Sincerely, Alana -------------- next part -------------- An embedded message was scrubbed... From: Gordon K Smyth <smyth@wehi.edu.au> Subject: [BioC] getting normexp (and log2 transformed) values - not further-processed M values Date: Wed, 7 Oct 2009 09:25:42 +1100 (AUS Eastern Daylight Time) Size: 5003 URL: <https: stat.ethz.ch="" pipermail="" bioconductor="" attachments="" 20091007="" ae4b6a5c="" attachment.eml="">