Display of histogram of RMA-processed data
3
0
Entering edit mode
@richard-friedman-513
Last seen 7.7 years ago
Dear Bioconductor Users, I wish to view the histogram of RMA processed data to see if the RMA algorithm converged histograms of a number of slides and got rid of a bothersome second peak. When I apply the command > hist(howard1.norm) where howard1.norm is the output of the RMA command, > howard1.norm <-rma(howard1.raw) I get the following errormessage: Error in hist.default(howard1.norm) : `x' must be numeric Jim MacDonald pointed out to me that a file made with the command > dat1 <- normalize.AffyBatch.quantiles(dat) can have its histogram displayed. However I want to be able to see the histogram of dat processed by rma. Is there any command that I can use to display the histogram of data processed by the rma command? Thanks and best wishes, Rich ------------------------------------------------------------ Richard A. Friedman, PhD Associate Research Scientist Herbert Irving Comprehensive Cancer Center Oncoinformatics Core Lecturer Department of Biomedical Informatics Box 95, Room 130BB or P&S 1-420C Columbia University Medical Center 630 W. 168th St. New York, NY 10032 (212)305-6901 (5-6901) (voice) friedman@cancercenter.columbia.edu http://cancercenter.columbia.edu/~friedman/ "I'm giving up coffee. Now that it looks like I might top 6 feet, I don't want to risk spoiling it." - Isaac Friedman, age 14 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1732 bytes Desc: not available Url : https://www.stat.math.ethz.ch/pipermail/bioconductor/attachments /20040513/034b0cfc/attachment.bin
Cancer Cancer • 955 views
ADD COMMENT
0
Entering edit mode
@james-w-macdonald-5106
Last seen 2 days ago
United States
Try plotDensity(exprs(howard1.norm)) Jim James W. MacDonald Affymetrix and cDNA Microarray Core University of Michigan Cancer Center 1500 E. Medical Center Drive 7410 CCGC Ann Arbor MI 48109 734-647-5623 >>> Richard Friedman <friedman@cancercenter.columbia.edu> 05/13/04 10:02AM >>> Dear Bioconductor Users, I wish to view the histogram of RMA processed data to see if the RMA algorithm converged histograms of a number of slides and got rid of a bothersome second peak. When I apply the command > hist(howard1.norm) where howard1.norm is the output of the RMA command, > howard1.norm <-rma(howard1.raw) I get the following errormessage: Error in hist.default(howard1.norm) : `x' must be numeric Jim MacDonald pointed out to me that a file made with the command > dat1 <- normalize.AffyBatch.quantiles(dat) can have its histogram displayed. However I want to be able to see the histogram of dat processed by rma. Is there any command that I can use to display the histogram of data processed by the rma command? Thanks and best wishes, Rich ------------------------------------------------------------ Richard A. Friedman, PhD Associate Research Scientist Herbert Irving Comprehensive Cancer Center Oncoinformatics Core Lecturer Department of Biomedical Informatics Box 95, Room 130BB or P&S 1-420C Columbia University Medical Center 630 W. 168th St. New York, NY 10032 (212)305-6901 (5-6901) (voice) friedman@cancercenter.columbia.edu http://cancercenter.columbia.edu/~friedman/ "I'm giving up coffee. Now that it looks like I might top 6 feet, I don't want to risk spoiling it." - Isaac Friedman, age 14
ADD COMMENT
0
Entering edit mode
@richard-friedman-513
Last seen 7.7 years ago
Skipped content of type multipart/alternative-------------- next part -------------- A non-text attachment was scrubbed... Name: howard1.raw.hist.pdf Type: application/pdf Size: 82927 bytes Desc: not available Url : https://www.stat.math.ethz.ch/pipermail/bioconductor/attachments /20040513/0713c419/howard1.raw.hist.pdf -------------- next part -------------- Skipped content of type multipart/alternative-------------- next part -------------- A non-text attachment was scrubbed... Name: howard.norm2.hist.pdf Type: application/pdf Size: 62205 bytes Desc: not available Url : https://www.stat.math.ethz.ch/pipermail/bioconductor/attachments /20040513/0713c419/howard.norm2.hist.pdf -------------- next part -------------- Skipped content of type multipart/alternative-------------- next part -------------- A non-text attachment was scrubbed... Name: howard1.norm.plotdensity.pdf Type: application/pdf Size: 86355 bytes Desc: not available Url : https://www.stat.math.ethz.ch/pipermail/bioconductor/attachments /20040513/0713c419/howard1.norm.plotdensity.pdf -------------- next part -------------- Skipped content of type multipart/alternative-------------- next part -------------- A non-text attachment was scrubbed... Name: howard1,raw.plotDensity.pdf Type: application/pdf Size: 73852 bytes Desc: not available Url : https://www.stat.math.ethz.ch/pipermail/bioconductor/attachments /20040513/0713c419/howard1raw.plotDensity.pdf -------------- next part -------------- Skipped content of type multipart/alternative
ADD COMMENT
0
Entering edit mode
Hello all I have seen such post RMA hist plots in my data too (human tumours on U133A chips). I did not see a nice smooth curve but curves that looked very similar to yours (two peaks etc). I think there is nothing wrong with the data, it probably has to do with the way you plot it. Indeed, if you use plot(density(exprs(rma), bw="nrd0", adjust=1)) instead and play around with the adjust value you will see that you can actually smoothen the way the data looks in the plots. This actually alters the width used for plotting I believe (?). I would be interested to hear what everybody thinks about this. Yours kindly, Lawrence _______________________________ Lawrence-Paul Petalidis Ph.D Candidate Department of Pathology Division of Molecular Histopathology University of Cambridge _______________________________ -----Original Message----- From: bioconductor-bounces@stat.math.ethz.ch [mailto:bioconductor-bounces@stat.math.ethz.ch] On Behalf Of Richard Friedman Sent: 13 May 2004 16:21 To: James MacDonald Cc: bioconductor@stat.math.ethz.ch Subject: Re: [BioC] Display of histogram of RMA-processed data On May 13, 2004, at 10:35 AM, James MacDonald wrote: Try plotDensity(exprs(howard1.norm)) Jim Jim (and everybody), Jim, Thanks for your advice. I am somewhat puzzled at the outcome. Here is is the hist plot of the unnormalized data: hist(howard1.raw): << File: howard1.raw.hist.pdf >> << File: howard.norm2.hist.pdf >> << File: howard1.norm.plotdensity.pdf >> << File: howard1,raw.plotDensity.pdf >> [[alternative HTML version deleted]]
ADD REPLY
0
Entering edit mode
if the concern is the last little bump for high concentrations, that is most likely due to saturation. this is usually due to some scanner tunning parameter being too high. no preprocession algorithm will (or should) get rid of that. On Thu, 13 May 2004, Richard Friedman wrote: > On May 13, 2004, at 10:35 AM, James MacDonald wrote: > > > Try plotDensity(exprs(howard1.norm)) > > > > Jim > > > > Jim (and everybody), > Jim, Thanks for your advice. > I am somewhat puzzled at the outcome. Here is is the hist plot of the > unnormalized data: > > hist(howard1.raw): >
ADD REPLY
0
Entering edit mode
@richard-friedman-513
Last seen 7.7 years ago
Ben, Here is the plot without the background correction: -------------- next part -------------- A non-text attachment was scrubbed... Name: howard1.norm4.nbg.pdf Type: application/pdf Size: 84406 bytes Desc: not available Url : https://www.stat.math.ethz.ch/pipermail/bioconductor/attachments /20040513/f0f01034/howard1.norm4.nbg.pdf -------------- next part -------------- Do you advise me not to use the background correction in this case? Thanks and best wishes, Rich On May 13, 2004, at 11:32 AM, Ben Bolstad wrote: > One feature that is pretty striking is that you have a reasonable > amount > of saturation in your data (this is why you have the bump at the high > end). > > The weird biomodality (ignoring the small bump at the high end) is an > artifact of the background correction. Try this to see that disappear. > > howard1.norm <- rma(howard1.raw,background=FALSE) > plotDensity(exprs(howard1.norm)) > > > Ben > > > On Thu, 2004-05-13 at 08:20, Richard Friedman wrote: >> On May 13, 2004, at 10:35 AM, James MacDonald wrote: >> >> Try plotDensity(exprs(howard1.norm)) >> >> Jim >> >> >> >> Jim (and everybody), >> Jim, Thanks for your advice. >> I am somewhat puzzled at the outcome. Here is is the hist plot of >> theunnormalized data: >> >> hist(howard1.raw): >> >> >> ______________________________________________________________________ >> >> >> Here is the hist plot of the data normailzed by quanitles alone >> howard1.norm2<-normalize.AffyBatch.quantiles(howard1.raw) >> hist(howard1.norm2) >> >> >> ______________________________________________________________________ >> >> >> >> All of the plots were superimposed. However, i wanted to see >> thedistribution from >> RMA. So I followed your instructions. >> To the file howard1.norm that was created in the following way: >> >> howard1.norm <- rma(howard1.raw) >> I used the plotDensty command >> >> plotDensity(exprs(howard1.norm)) >> >> and got >> >> >> >> >> James W. MacDonald >> Affymetrix and cDNA Microarray Core >> >> >> ______________________________________________________________________ >> >> >> Shouldn't the distrubution be unimodal as in when just quantiles >> wasapplied? Also, is this this distrbution >> anything to worry about. >> >> Also, >> plotDensity(exprs(object))does not seem to be the equivalent of >> his(object). when I applied it to my raw dataset I got: >> >> >> ______________________________________________________________________ >> >> So, plotDensity doesn;t seem to be theequivalent of hist. >> Also, does the bimodal nature of the plotDensity of the rma plot >> pose a problem. It is trimodla if you count the blip at the end. >> >> Thanks and best wishes, >> Rich >> >> >> University of Michigan Cancer Center >> 1500 E. Medical Center Drive >> 7410 CCGC >> Ann Arbor MI 48109 >> 734-647-5623 >> >> Richard >> >> Friedman<friedman@cancercenter.columbia.edu> 05/13/04 >> 10:02AM >>> >> Dear Bioconductor Users, >> >> I wish to view the histogram of RMA processed data to >> see if the RMA algorithm converged histograms of a number of >> slides >> and got rid of a bothersome second peak. When I apply the >> command >> >> hist(howard1.norm) >> where howard1.norm is the output of the RMA command, >> >> howard1.norm <-rma(howard1.raw) >> >> I get the following errormessage: >> Error in hist.default(howard1.norm) : `x' must be numeric >> >> Jim MacDonald pointed out to me that a file made with the >> command >> >> dat1 <- normalize.AffyBatch.quantiles(dat) >> >> can have its histogram displayed. However I want to be able to >> see the >> histogram of dat processed by rma. >> >> Is there any command that I can use to display the histogram >> of data >> processed by the rma command? >> >> Thanks and best wishes, >> Rich >> ------------------------------------------------------------ >> Richard A. Friedman, PhD >> Associate Research Scientist >> Herbert Irving Comprehensive Cancer Center >> Oncoinformatics Core >> Lecturer >> Department of Biomedical Informatics >> Box 95, Room 130BB or P&S 1-420C >> Columbia University Medical Center >> 630 W. 168th St. >> New York, NY 10032 >> (212)305-6901 (5-6901) (voice) >> friedman@cancercenter.columbia.edu >> http://cancercenter.columbia.edu/~friedman/ >> >> "I'm giving up coffee. Now that it looks like I might top >> 6 feet, I don't want to risk spoiling it." >> - Isaac Friedman, age 14 >> >> ------------------------------------------------------------ >> Richard A. Friedman, PhD >> Associate Research Scientist >> Herbert Irving Comprehensive Cancer Center >> Oncoinformatics Core >> Lecturer >> Department of Biomedical Informatics >> Box 95, Room 130BB or P&S 1-420C >> Columbia University Medical Center >> 630 W. 168th St. >> New York, NY 10032 >> (212)305-6901 (5-6901) (voice) >> friedman@cancercenter.columbia.edu >> http://cancercenter.columbia.edu/~friedman/ >> >> "I'm giving up coffee. Now that it looks like I might top >> 6 feet, I don't want to risk spoiling it." >> - Isaac Friedman, age 14 >> >> >> >> ______________________________________________________________________ >> _______________________________________________ >> Bioconductor mailing list >> Bioconductor@stat.math.ethz.ch >> https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor > -- > Ben Bolstad <bolstad@stat.berkeley.edu> > http://www.stat.berkeley.edu/~bolstad > > ------------------------------------------------------------ Richard A. Friedman, PhD Associate Research Scientist Herbert Irving Comprehensive Cancer Center Oncoinformatics Core Lecturer Department of Biomedical Informatics Box 95, Room 130BB or P&S 1-420C Columbia University Medical Center 630 W. 168th St. New York, NY 10032 (212)305-6901 (5-6901) (voice) friedman@cancercenter.columbia.edu http://cancercenter.columbia.edu/~friedman/ "I'm giving up coffee. Now that it looks like I might top 6 feet, I don't want to risk spoiling it." - Isaac Friedman, age 14
ADD COMMENT

Login before adding your answer.

Traffic: 139 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6