[R] how to partitioning a clustered heatmap

[Fededa Juan Pablo]

Hello all,

I'm doing heatmap.2 graphics of matrix x with hclust clustering in the rows, as an argument in heatmap.2 .
The problem is that the heatmap contains 1000 genes, and I want to cut it in several pieces, but after the clustering.
If I use x[1:100,] as delimiting argument for the first 100 clustered rows(genes) in the matrix, it doesn't work, it takes the 100 first rows in the matrix and does the clustering over them, and not first the clustering over the 1000 genes and then graph the first 100 rows of the clustered matrix.
The entire argument of heatmap I'm using is something like this:

heatmap.2(x, scale="none", cexRow=0.00001, keysize=0.8, hclustfun = hclust, sepcolor="none",Colv=FALSE, dendrogram="row", trace=c("none"),density.info=c("none"))

If you can help me, will be great.
Thanks in advance,


juan



Juan Pablo Fededa
Institute of Biochemistry-ETH Zurich
HPM,D11.1
Schafmattstrasse 18
CH-8093, Zurich
Switzerland
+41 44 633 60 26  phone
+41 44 632 12 69 fax
jpfededa at bc.biol.ethz.ch
www.lmc.ethz.ch/IBC/