[BioC] Methods for building heatmap for Illumina gene expression data

Sean Davis sdavis2 at mail.nih.gov
Fri May 23 14:03:38 CEST 2008

On Thu, May 22, 2008 at 7:05 PM, ss <affysnp at gmail.com> wrote:
> Hi Sean,
> Thanks for pointing it out. The reason that I would like to supply with a
> txt
> file of Illumina gene expression data is that I find it hard to use either
> beadarray
> or lumi to subset gene expression data based on genes of your interest.
> Please
> correct me if I am wrong. Anyway, I will check it out.

I do not find it particularly difficult to subset genes and samples in
either of those packages, and one can always use the exprs() method
for each to extract a raw matrix and subset that.  For the beadarray
package, see the vignette entitled "beadsummary.pdf", paying
particular attention to the "BSData" section.  Similar methods work
with the lumi package.

If by "find it hard to use either beadarray or lumi to subset..." you
mean that you do not know where to start, I have given a hint above.
Also, many of the data classes used in Bioconductor are based on the
ExpressionSet class (or the eSet class).  There is extensive
documentation in the Biobase package for taking subsets of data that
will likely apply to both beadarray and lumi subsetting.  It is
worthwhile working through the examples in the various vignettes and
reading help pages if you would like to maximize your productivity
with bioconductor.

Hope that helps get you going.


> On Thu, May 22, 2008 at 4:29 PM, Sean Davis <sdavis2 at mail.nih.gov> wrote:
>> Hi, Allen.  As for illumina data, I would look into using the
>> beadarray or lumi packages.  No reason to use excel for these types of
>> things.
>> As for heatmap, try searching the mailing list here:
>> http://dir.gmane.org/gmane.science.biology.informatics.conductor
>> If you type in the keywords heatmap and color, you will see numerous
>> potential possibilities.
>> Sean

More information about the Bioconductor mailing list