[BioC] recipe for heatmap

[Sean Davis]

Marc,

The heatmap function does the clustering from the data.  In other words, you
call heatmap with your matrix of gene expression values and heatmap clusters
the samples and genes.  You can optionally ask heatmap to not cluster either
rows (genes) or columns (samples).


Sean

On 7/2/04 9:45 AM, "Marc Salit" <salit at nist.gov> wrote:

> Hi all --
> 
> I've data from 12 affy arrays, 5 in one class, 7 in the other. I am
> looking for genes that classify a sample into one or the other classes.
> 
> One thing I'd like to do is to be able to visually evaluate the power
> that exists to distinguish the classes, using clustered heatmaps. I've
> done some of this in the dchip program and NCI's BRB arraytools, but I
> haven't found an appropriate vignette or lab/tutorial example in the R
> canon.
> 
> I am guessing that I need to filter the genes to those that are
> differentially expressed, then use the 'heatmap' function. One thing I
> can't see immediately is how to identify the different classes in the
> call to heatmap...
> 
> I'm certain that I'm not the first to tackle this, can someone share
> their experience?
> 
> Thanks in advance!
> Marc