[R] HELP ON SCALING GENE EXPRESSION DATA TO -1,0,1
mtmorgan at fhcrc.org
Mon Nov 17 16:57:44 CET 2008
"paul murima" <paul.murima at gmail.com> writes:
> Hello R-Community,
> I am a rookie in R and I am fascinated with the power of bio computing
> by R. I am analysing gene expression data from Real time PCR. I have
> used absolute gene quantitation to measure gene copy number in all my
> transcripts. All my data has been normalised them to a housekeeping
> gene, which is constitutive expressed.
> My problem is as follows.
> After normalising some of the genes (which are feature vectors in my
> matrix of 26 treatment conditions) are up regulated to over 50 fold.
Do you know about http://bioconductor.org ? Reviewing their mailing
list archives and asking on their mailing list will likely lead to a
number of informed opinions about normalization.
> Now, when plot my heatmap.2 of the matrix, almost everything shows up
> as red, in a green-red colour scale.
> i.e. from -1,0,1. Is there a way to scale all the data on the matrix
> such that when i plot it on a heat map it it reflecting the relative
> expression of of each feature, against the treatment condition without
> being fixed to the -1, to 1 scale.
> Below is how the code i used.
> zcz<- read.table("chitekete.csv", header = T, row.names=1, sep=",");
> xcc <- cor(zcz);
> pdf("SIGA.pdf", height=10,width=9)
> heatmap.2(xcc, margins= c(9,9), col = bluered(64), trace=c("none"),
> breaks=c(seq(-1,1,1/32)), symkey=TRUE, density.info="histogram",
> I thank you all.
> Paul Murima
> R-help at r-project.org mailing list
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
Computational Biology / Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N.
PO Box 19024 Seattle, WA 98109
Location: Arnold Building M2 B169
Phone: (206) 667-2793
More information about the R-help