heatmap.2
5
0
Entering edit mode
Benjamin Otto ▴ 830
@benjamin-otto-1519
Last seen 8.2 years ago
How about data standardization before plotting? Should center the values automatically around 0. There is some flag in the function to da that, but I don't have it in mind right now. regards, Benjamin Am Dienstag, den 31.07.2007, 09:56 +0200 schrieb Kauer Maximilian Otto: > Dear list, > > Is there a way to force the colors in heatmap.2 to center on 0 i.e to to force 0 to be white even though the spread in the data is unequal (say -7 to +4)? I'm using bluered, but could use different color sets. > > Thank You, > Max > > _______________________________________________ > 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 > -- Pflichtangaben gem?? Gesetz ?ber elektronische Handelsregister und Genossenschaftsregister sowie das Unternehmensregister (EHUG): Universit?tsklinikum Hamburg-Eppendorf K?rperschaft des ?ffentlichen Rechts Gerichtsstand: Hamburg Vorstandsmitglieder: Prof. Dr. J?rg F. Debatin (Vorsitzender) Dr. Alexander Kirstein Ricarda Klein Prof. Dr. Dr. Uwe Koch-Gromus
• 1.6k views
ADD COMMENT
0
Entering edit mode
Jenny Drnevich ★ 2.2k
@jenny-drnevich-382
Last seen 8.2 years ago
Hi, I prefer not to standardize around 0 or use the scale="row" option of heatmap2, because it's not always clear what the resulting values mean. What I've done is to add a row to the data to force the center to be on zero: > class(mean.sig.05) [1] "matrix" > dim(mean.sig.05) [1] 101 15 > max(mean.sig.05) [1] 3.24023 > min(mean.sig.05) [1] -3.120530 > mean.sig.05 <- rbind(mean.sig.05,fake=c(3.24,rep(0,13),-3.24)) > rownames(mean.sig.05) <- NULL >hmcol2 <- colorRampPalette(c("red","white","blue"))(256) > heatmap_2(mean.sig.05,col=hmcol2,legend=1,do.dendro=c(T,T),scale="n one") Of course, it's not ideal to have a faked row of data in the plot, but it doesn't really show if you have >50 genes. Cheers, Jenny At 06:24 AM 8/1/2007, Benjamin Otto wrote: >How about data standardization before plotting? Should center the values >automatically around 0. There is some flag in the function to da that, >but I don't have it in mind right now. > >regards, > >Benjamin > > >Am Dienstag, den 31.07.2007, 09:56 +0200 schrieb Kauer Maximilian Otto: > > Dear list, > > > > Is there a way to force the colors in > heatmap.2 to center on 0 i.e to to force 0 to > be white even though the spread in the data is > unequal (say -7 to +4)? I'm using bluered, but could use different color sets. > > > > Thank You, > > Max > > > > _______________________________________________ > > 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 > > > > > >-- >Pflichtangaben gem?? Gesetz ?ber elektronische >Handelsregister und Genossenschaftsregister >sowie das Unternehmensregister (EHUG): > >Universit?tsklinikum Hamburg-Eppendorf >K?rperschaft des ?ffentlichen Rechts >Gerichtsstand: Hamburg > >Vorstandsmitglieder: >Prof. Dr. J?rg F. Debatin (Vorsitzender) >Dr. Alexander Kirstein >Ricarda Klein >Prof. Dr. Dr. Uwe Koch-Gromus > > >_______________________________________________ >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 Jenny Drnevich, Ph.D. Functional Genomics Bioinformatics Specialist W.M. Keck Center for Comparative and Functional Genomics Roy J. Carver Biotechnology Center University of Illinois, Urbana-Champaign 330 ERML 1201 W. Gregory Dr. Urbana, IL 61801 USA ph: 217-244-7355 fax: 217-265-5066 e-mail: drnevich at uiuc.edu
ADD COMMENT
0
Entering edit mode
An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20070801/ 28f2be10/attachment.pl
ADD REPLY
0
Entering edit mode
Jenny Drnevich ★ 2.2k
@jenny-drnevich-382
Last seen 8.2 years ago
An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20070801/ a85df863/attachment.pl
ADD COMMENT
0
Entering edit mode
Jenny Drnevich ★ 2.2k
@jenny-drnevich-382
Last seen 8.2 years ago
An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20070801/ 996f80f3/attachment.pl
ADD COMMENT
0
Entering edit mode
An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20070801/ 93038dc3/attachment.pl
ADD REPLY
0
Entering edit mode
An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20070801/ 34e30db8/attachment.pl
ADD REPLY
0
Entering edit mode
An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20070801/ e4d33c99/attachment.pl
ADD REPLY
0
Entering edit mode
At 02:10 PM 8/1/2007, carol white wrote: >with greenred, I get red and black and with redgreen, I get green >and black. I wonder if I could have the same colors as TreeView or MapleTree. Hi Carol, I wouldn't say your inability to get red-black-green in your plots is a failure of the color schemes, but rather that your expression differences are almost all in one direction. You should take a careful look at this, and the data values that you are plotting. Maybe you've done an unusual transformation somewhere along the way? If the expression differences are truly almost all in one direction, that's a warning flag that you should think about how you pre-processed your data and the assumptions behind the various methods. For example, all expression changes in one direction would mean that you wouldn't expect the M values in an MA plot to be centered on zero. I would take a very careful look at my data at all steps in the process to see if any mistakes have been made, or if the normalization steps were actually appropriate for this data set. Jenny >Artur Veloso <abveloso at="" gmail.com=""> wrote: Hi Carol, > >on the library "gplots" there is a function called "greenred" that >should do what I think you want to do. >Use col=greenred(50) and see if it works out fine. > >Cheers, > >Artur > >On 8/1/07, carol white <wht_crl at="" yahoo.com=""> wrote: >with col=colorRampPalette(c("red","black","green"))(256), green is >replaced by black and red by green, hence, red doesn't appear any more > At 01:18 PM 8/1/2007, carol white wrote: > thanks for your quick reply. > > I am looking to get something closer to what is in gene expression > papers like http://rana.lbl.gov/papers/Eisen_PNAS_1998.pdf even > though that they use other clustering tools like TreeView. If I > could get almost the same red and green, it woudl be good. in this > heatmaps, black shows missing values or no gene expression. So I > don't know how to integrate this. > Try col=colorRampPalette(c("red","black","green"))(256) > > Best, > Jenny > > > > Best > > Jenny Drnevich <drnevich at="" uiuc.edu=""> wrote: > > Hi Carol, > > > R has 657 colors by name (see colors() ), but I don't know if > there is an easy way to look through them to see which of the red > and green options give the hues you want. However, I would caution > you against using a red/green scheme. I know it's used because the > dye colors for 2-color arrays are red and green, but for display > purposes red/green is a BAD choice because 10% of males are > red-green color blind. > > > Jenny > > > At 12:42 PM 8/1/2007, carol white wrote: > > Hi, > > Regarding heatmap, I would like to display the colors in green > and red. Is setting the parameter col in heatmap to > col=colorRampPalette(c("red","green"))(256) the best solution? > green seems to be tense and red seems to be dull (see attached). > > > > Another solution would have been to use brewer.pal in > RColorBrewer package but I didn't find any name with only red and > green option for brewer.pal. > > > any suggestion is welcome. > > > thanks > > > carol > > > > Building a website is a piece of cake. > > > Jenny Drnevich, Ph.D. > > > Functional Genomics Bioinformatics Specialist > > W.M. Keck Center for Comparative and Functional Genomics > > Roy J. Carver Biotechnology Center > > University of Illinois, Urbana-Champaign > > > 330 ERML > > 1201 W. Gregory Dr. > > Urbana, IL 61801 > > USA > > > ph: 217-244-7355 > > fax: 217-265-5066 > > e-mail: drnevich at uiuc.edu > > > > Looking for a deal? Find great prices on flights and hotels with > Yahoo! FareChase. > > >--------------------------------- >Pinpoint customers who are looking for what you sell. > [[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 > > > > > >--------------------------------- >Pinpoint customers who are looking for what you sell. > [[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 Jenny Drnevich, Ph.D. Functional Genomics Bioinformatics Specialist W.M. Keck Center for Comparative and Functional Genomics Roy J. Carver Biotechnology Center University of Illinois, Urbana-Champaign 330 ERML 1201 W. Gregory Dr. Urbana, IL 61801 USA ph: 217-244-7355 fax: 217-265-5066 e-mail: drnevich at uiuc.edu
ADD REPLY
0
Entering edit mode
An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20070801/ d22b694a/attachment.pl
ADD REPLY
0
Entering edit mode
An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20070801/ 5dc00b75/attachment.pl
ADD REPLY
0
Entering edit mode
An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20070801/ 29283c64/attachment.pl
ADD REPLY
0
Entering edit mode
Matthias Kohl ▴ 160
@matthias-kohl-1678
Last seen 8.2 years ago
Hi, I implemented a little function in my R package SLmisc which can help to adjust the colors; see library(SLmisc) example(heatmapCol) Best, Matthias ----- original message -------- Subject: Re: [BioC] heatmap.2 Sent: Wed, 01 Aug 2007 From: carol white<wht_crl@yahoo.com> > back to standardization discussion, is it better to do > 1- mean centering, or > 2- median centering of variables or > 3- mean = 0, standard deviation = 1 of variables? > > 1 and 2 could be done with scale function (center = T). so we would know > what the results would be. > > Jenny Drnevich <drnevich at="" uiuc.edu=""> wrote: Hi, > > I prefer not to standardize around 0 or use the > scale="row" option of heatmap2, because it's not > always clear what the resulting values mean. What > I've done is to add a row to the data to force the center to be on zero: > > > class(mean.sig.05) > [1] "matrix" > > > dim(mean.sig.05) > [1] 101 15 > > > max(mean.sig.05) > [1] 3.24023 > > > min(mean.sig.05) > [1] -3.120530 > > > mean.sig.05 <- rbind(mean.sig.05,fake=c(3.24,rep(0,13),-3.24)) > > > rownames(mean.sig.05) <- NULL > > >hmcol2 <- colorRampPalette(c("red","white","blue"))(256) > > > heatmap_2(mean.sig.05,col=hmcol2,legend=1,do.dendro=c(T,T),scale= "none") > > Of course, it's not ideal to have a faked row of > data in the plot, but it doesn't really show if you have >50 genes. > > Cheers, > Jenny > > At 06:24 AM 8/1/2007, Benjamin Otto wrote: > >How about data standardization before plotting? Should center the values > >automatically around 0. There is some flag in the function to da that, > >but I don't have it in mind right now. > > > >regards, > > > >Benjamin > > > > > >Am Dienstag, den 31.07.2007, 09:56 +0200 schrieb Kauer Maximilian Otto: > > > Dear list, > > > > > > Is there a way to force the colors in > > heatmap.2 to center on 0 i.e to to force 0 to > > be white even though the spread in the data is > > unequal (say -7 to +4)? I'm using bluered, but could use different color > sets. > > > > > > Thank You, > > > Max > > > > > > _______________________________________________ > > > 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 > > > > > > > > > > >-- > >Pflichtangaben gem?? Gesetz ?ber elektronische > >Handelsregister und Genossenschaftsregister > >sowie das Unternehmensregister (EHUG): > > > >Universit?tsklinikum Hamburg-Eppendorf > >K?rperschaft des ?ffentlichen Rechts > >Gerichtsstand: Hamburg > > > >Vorstandsmitglieder: > >Prof. Dr. J?rg F. Debatin (Vorsitzender) > >Dr. Alexander Kirstein > >Ricarda Klein > >Prof. Dr. Dr. Uwe Koch-Gromus > > > > > >_______________________________________________ > >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 > > Jenny Drnevich, Ph.D. > > Functional Genomics Bioinformatics Specialist > W.M. Keck Center for Comparative and Functional Genomics > Roy J. Carver Biotechnology Center > University of Illinois, Urbana-Champaign > > 330 ERML > 1201 W. Gregory Dr. > Urbana, IL 61801 > USA > > ph: 217-244-7355 > fax: 217-265-5066 > e-mail: drnevich at uiuc.edu > > _______________________________________________ > 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 > > > > Jenny Drnevich <drnevich at="" uiuc.edu=""> wrote: Hi, > > I prefer not to standardize around 0 or use the > scale="row" option of heatmap2, because it's not > always clear what the resulting values mean. What > I've done is to add a row to the data to force the center to be on zero: > > > class(mean.sig.05) > [1] "matrix" > > > dim(mean.sig.05) > [1] 101 15 > > > max(mean.sig.05) > [1] 3.24023 > > > min(mean.sig.05) > [1] -3.120530 > > > mean.sig.05 <- rbind(mean.sig.05,fake=c(3.24,rep(0,13),-3.24)) > > > rownames(mean.sig.05) <- NULL > > >hmcol2 <- colorRampPalette(c("red","white","blue"))(256) > > > heatmap_2(mean.sig.05,col=hmcol2,legend=1,do.dendro=c(T,T),scale= "none") > > Of course, it's not ideal to have a faked row of > data in the plot, but it doesn't really show if you have >50 genes. > > Cheers, > Jenny > > At 06:24 AM 8/1/2007, Benjamin Otto wrote: > >How about data standardization before plotting? Should center the values > >automatically around 0. There is some flag in the function to da that, > >but I don't have it in mind right now. > > > >regards, > > > >Benjamin > > > > > >Am Dienstag, den 31.07.2007, 09:56 +0200 schrieb Kauer Maximilian Otto: > > > Dear list, > > > > > > Is there a way to force the colors in > > heatmap.2 to center on 0 i.e to to force 0 to > > be white even though the spread in the data is > > unequal (say -7 to +4)? I'm using bluered, but could use different color > sets. > > > > > > Thank You, > > > Max > > > > > > _______________________________________________ > > > 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 > > > > > > > > > > >-- > >Pflichtangaben gem?? Gesetz ?ber elektronische > >Handelsregister und Genossenschaftsregister > >sowie das Unternehmensregister (EHUG): > > > >Universit?tsklinikum Hamburg-Eppendorf > >K?rperschaft des ?ffentlichen Rechts > >Gerichtsstand: Hamburg > > > >Vorstandsmitglieder: > >Prof. Dr. J?rg F. Debatin (Vorsitzender) > >Dr. Alexander Kirstein > >Ricarda Klein > >Prof. Dr. Dr. Uwe Koch-Gromus > > > > > >_______________________________________________ > >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 > > Jenny Drnevich, Ph.D. > > Functional Genomics Bioinformatics Specialist > W.M. Keck Center for Comparative and Functional Genomics > Roy J. Carver Biotechnology Center > University of Illinois, Urbana-Champaign > > 330 ERML > 1201 W. Gregory Dr. > Urbana, IL 61801 > USA > > ph: 217-244-7355 > fax: 217-265-5066 > e-mail: drnevich at uiuc.edu > > _______________________________________________ > 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 > > > > --------------------------------- > Luggage? GPS? Comic books? > > [[alternative HTML version deleted]] > > --- original message end ----
ADD COMMENT
0
Entering edit mode
Max Kauer ▴ 140
@max-kauer-2254
Last seen 6.2 years ago
Hi again I have a (very simple) question about heatmap.2 (Thanks for all the answers the other day!). Has anybody been succeful in scaling the main title? cex.main does not work Thanks, Max
ADD COMMENT
0
Entering edit mode
An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20070810/ 3dea3b82/attachment.pl
ADD REPLY
0
Entering edit mode
An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/bioconductor/attachments/20070810/ 36ae0e97/attachment.pl
ADD REPLY

Login before adding your answer.

Traffic: 542 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