[BioC] heatmap.2 ..help..how to cluster column ..but not Rows..

Saurin D. Jani saurin_jani at yahoo.com
Fri Apr 15 14:50:02 CEST 2011


Hi Jason, Sean,

Thank you so much for your time and reply. I really appreciate that. Now, it does work. The only explanation I have it...when I fixed my strings for input sample names..I got it working..as Jason showed. 


have a fantastic weekend,
Thanks,
Saurin

--- On Fri, 4/15/11, Jason Shoemaker <jshoe at ims.u-tokyo.ac.jp> wrote:

> From: Jason Shoemaker <jshoe at ims.u-tokyo.ac.jp>
> Subject: Re: [BioC] heatmap.2 ..help..how to cluster column ..but not Rows..
> To: "saurin_jani" <saurin_jani at yahoo.com>
> Cc: "Sean Davis" <sdavis2 at mail.nih.gov>, "Bioconductor Bioconductor" <bioconductor at stat.math.ethz.ch>
> Date: Friday, April 15, 2011, 12:42 AM
> hmm. I am not sure what else to try
> then. I followed that example, ran
> this code:
> > hclust2 <- function(x, method="average", ...)
> +   hclust(x, method=method, ...)
> > dist2 <- function(x, ...)
> +   as.dist(1-cor(t(x), method="pearson"))
> heatmap.2(Rowv=F,as.matrix(mtcars), distfun=dist2,
> hclustfun=hclust2)
> 
> it seems to run fine...I attached the figure here.
> Hopefully someone
> else knows the solution.
> Goodluck!
> Jason
> 
> 
> 
> 
> On Fri, Apr 15, 2011 at 12:12 PM, Saurin D. Jani <saurin_jani at yahoo.com>
> wrote:
> > Hi Jason,
> >
> > Thanks. but, somehow that does not work..!
> >
> > hclust2 <- function(x, method="average")
> > dist2 <- function(x)
> >  as.dist(1-cor(t(x), method="pearson"))
> >
> > heatmap.2(FeatureX, distfun=dist2,
> hclustfun=hclust2);
> >
> > Error in UseMethod("as.dendrogram") :
> >  no applicable method for 'as.dendrogram' applied to
> an object of class "function"
> >
> >
> > ## one small experiment:
> >
> > heatmap.2(FeatureX,col=bluered,Colv =
> as.dendrogram(hclust(col.dist, method = "centroid")),
> dendrogram = "column", scale = "none",key = TRUE, Rowv =
> FALSE, density.info = "none",tracecol="black");
> >
> >
> > I suspect that...my column names are bit different in
> length...when I tried with "AA_0001" and so on..above
> heatmap.2 call works exactly what I want...!!
> >
> > Did any one experienced similar ? this is just me..!
> >
> > Thanks,
> > Saurin
> >
> > --- On Thu, 4/14/11, Jason Shoemaker <jshoe at ims.u-tokyo.ac.jp>
> wrote:
> >
> >> From: Jason Shoemaker <jshoe at ims.u-tokyo.ac.jp>
> >> Subject: Re: [BioC] heatmap.2 ..help..how to
> cluster column ..but not Rows..
> >> To: "saurin_jani" <saurin_jani at yahoo.com>
> >> Cc: "Sean Davis" <sdavis2 at mail.nih.gov>,
> "Bioconductor Bioconductor" <bioconductor at stat.math.ethz.ch>
> >> Date: Thursday, April 14, 2011, 11:01 PM
> >> This previous discussion seems to
> >> answer the probleb:
> >> https://stat.ethz.ch/pipermail/bioconductor/2009-July/028898.html
> >> Jason
> >>
> >> On Fri, Apr 15, 2011 at 11:48 AM, Saurin D. Jani
> <saurin_jani at yahoo.com>
> >> wrote:
> >> > Hi Sean,
> >> >
> >> > Thanks for your reply. I tried all
> those..options in
> >> arguments. When I do Rowv = FALSE.. column
> clustering is not
> >> there.
> >> >
> >> > thats the reason I got confused..!!
> >> >
> >> > heatmap.2(FeatureX,col=bluered,Colv =
> >> as.dendrogram(hclust(col.dist, method =
> "centroid")),
> >> dendrogram = "column", scale = "none",key = TRUE,
> Rowv =
> >> FALSE, density.info = "none",tracecol="black");
> >> >
> >> >
> >> > Thanks,
> >> > Saurin
> >> >
> >> >
> >> > --- On Thu, 4/14/11, Sean Davis <sdavis2 at mail.nih.gov>
> >> wrote:
> >> >
> >> >> From: Sean Davis <sdavis2 at mail.nih.gov>
> >> >> Subject: Re: [BioC] heatmap.2 ..help..how
> to
> >> cluster column ..but not Rows..
> >> >> To: saurin_jani at yahoo.com
> >> >> Cc: "Bioconductor Bioconductor" <bioconductor at stat.math.ethz.ch>
> >> >> Date: Thursday, April 14, 2011, 10:45 PM
> >> >> On Thu, Apr 14, 2011 at 10:34 PM,
> >> >> Saurin D. Jani <saurin_jani at yahoo.com>
> >> >> wrote:
> >> >> > Hi Sean,
> >> >> >
> >> >> > heatmap.2(FeatureX,col=bluered,Colv
> =
> >> >> as.dendrogram(hclust(col.dist, method =
> >> "centroid")), scale
> >> >> = "none",key = TRUE, dendrogram =
> "column",
> >> density.info =
> >> >> "none",tracecol="black");
> >> >> >
> >> >> >
> >> >> > This one gives dendrogram on column
> and key
> >> is there
> >> >> as well..the only problem is Rows are
> being
> >> clustered as
> >> >> well...how do I prevent rows to get
> clustered?
> >> >> >
> >> >>
> >> >> Take a look at the Rowv argument;
> Rowv=FALSE is
> >> probably
> >> >> what you want.
> >> >>
> >> >> Sean
> >> >>
> >> >>
> >> >> > Thanks,
> >> >> > Saurin
> >> >> >
> >> >> >
> >> >> >
> >> >> > --- On Thu, 4/14/11, Sean Davis
> <sdavis2 at mail.nih.gov>
> >> >> wrote:
> >> >> >
> >> >> >> From: Sean Davis <sdavis2 at mail.nih.gov>
> >> >> >> Subject: Re: [BioC] heatmap.2
> ..help..how
> >> to
> >> >> cluster column ..but not Rows..
> >> >> >> To: saurin_jani at yahoo.com
> >> >> >> Cc: "Bioconductor Bioconductor"
> <bioconductor at stat.math.ethz.ch>
> >> >> >> Date: Thursday, April 14, 2011,
> 10:24 PM
> >> >> >> On Thu, Apr 14, 2011 at 9:49
> PM,
> >> >> >> Saurin D. Jani <saurin_jani at yahoo.com>
> >> >> >> wrote:
> >> >> >> > Hi BioC,
> >> >> >> >
> >> >> >> > This must be simple but
> somehow I
> >> can not be
> >> >> able to
> >> >> >> do it...
> >> >> >> > How can I cluster samples
> only..
> >> below code
> >> >> is giving
> >> >> >> me dendrogram on both rows and
> clumns...!
> >> if I do
> >> >> Rowv =
> >> >> >> FALSE..then I don't see any
> colors and
> >> KEY.
> >> >> >> >
> >> >> >>
> >> >> >> Hi, Saurin.
> >> >> >>
> >> >> >> See the "dendrogram" argument
> in
> >> heatmap.2 help.
> >> >> In
> >> >> >> particular,
> >> >> >> dendrogram="column" will turn
> off the
> >> row
> >> >> dendrogram.
> >> >> >>
> >> >> >> Sean
> >> >> >>
> >> >> >> > Please see this:
> >> >> >> >
> >> >> >> > FeatureX is Matrix with
> proper rows
> >> and
> >> >> colmns.
> >> >> >> >
> >> >> >> >
> >> >> >> > row.dist <- as.dist(1 -
> >> >> cor(t(FeatureX)));
> >> >> >> > col.dist <- as.dist(1 -
> >> cor(FeatureX));
> >> >> >> >
> heatmap.2(FeatureX,col=bluered,Colv
> >> =
> >> >> >> as.dendrogram(hclust(col.dist,
> method =
> >> >> "centroid")), scale
> >> >> >> = "none",key = TRUE,density.info
> =
> >> >> >> "none",tracecol="black");
> >> >> >> >
> >> >> >> > Thank you in advance,
> >> >> >> > Saurin
> >> >> >> >
> >> >> >> >
> >> >>
> _______________________________________________
> >> >> >> > Bioconductor mailing list
> >> >> >> > Bioconductor at r-project.org
> >> >> >> > https://stat.ethz.ch/mailman/listinfo/bioconductor
> >> >> >> > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor
> >> >> >> >
> >> >> >>
> >> >> >
> >> >> >
> >> _______________________________________________
> >> >> > Bioconductor mailing list
> >> >> > Bioconductor at r-project.org
> >> >> > https://stat.ethz.ch/mailman/listinfo/bioconductor
> >> >> > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor
> >> >> >
> >> >>
> >> >
> >> >
> _______________________________________________
> >> > Bioconductor mailing list
> >> > Bioconductor at r-project.org
> >> > https://stat.ethz.ch/mailman/listinfo/bioconductor
> >> > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor
> >> >
> >>
> >>
> >>
> >> --
> >> Jason Shoemaker, Ph D
> >> Japanese Science and Technology Agency
> >> Infection Induced Host Response Laboratory
> >> The Institute of Medical Science,
> >> University of Tokyo
> >>
> >
> 
> 
> 
> -- 
> Jason Shoemaker, Ph D
> Japanese Science and Technology Agency
> Infection Induced Host Response Laboratory
> The Institute of Medical Science,
> University of Tokyo
>



More information about the Bioconductor mailing list