[BioC] LIMMA - summarization of Illumina probes for same gene

Gordon K Smyth smyth at wehi.EDU.AU
Tue Mar 13 23:43:06 CET 2012

Dear Kripa,

Please see the help page


If you call avereps() with ID equal to a vector of gene symbols, then it 
will average by genes.

Best wishes

Professor Gordon K Smyth,
Bioinformatics Division,
Walter and Eliza Hall Institute of Medical Research,
1G Royal Parade, Parkville, Vic 3052, Australia.
Tel: (03) 9345 2326, Fax (03) 9347 0852,
smyth at wehi.edu.au

On Tue, 13 Mar 2012, Kripa R wrote:

> Yes, I see what you mean; each illumina probe provides unique 
> information. I guess all the affy summarization options threw me off 
> (avgdiff, liwong, mas, medianpolish, playerout) One thing, the avereps 
> function would be looking for identical ProbeID but would it be possible 
> to change this to look for identical gene names/ look for replicates 
> using a different column?
> Thanks
>> Date: Mon, 12 Mar 2012 09:09:15 +1100
>> From: smyth at wehi.EDU.AU
>> To: kripa777 at hotmail.com
>> CC: bioconductor at r-project.org
>> Subject: LIMMA - summarization of Illumina probes for same gene
>> Dear Kripa,
>> Short answer:
>> If you really must average over multiple probes for a given gene, then the
>> limma function you are looking for is avereps().
>> Longer answer:
>> This question has come up a few times recently on this mailing list, and I
>> wonder whether you and other posters might be thinking of summarization as
>> is done by the RMA algorithm for Affymetrix arrays over multiple probes
>> in a probe-set.  The equivalent of this for an Illumina BeadChip is to
>> summarize the multiple beads for each probe on each array.  This step is
>> done by the Illumina Genome Studio software, so has already been completed
>> before entry into limma.
>> On the other hand, summarizing multiple probes for a given gene (or
>> multiple probe-sets for an Affy array) is not recommended as a routine
>> procedure by any of the Bioconductor software developers.  The whole
>> reason that multiple probes are provided by Illumina for a given gene is
>> that the gene may take different isoforms to which the different probes
>> may respond.  Why would you want to lose this information by pre-emptively
>> averaging?
>> Best wishes
>> Gordon
>>> Date: Fri, 9 Mar 2012 21:42:58 +0000
>>> From: Kripa R <kripa777 at hotmail.com>
>>> To: <bioconductor at r-project.org>
>>> Subject: [BioC] LIMMA - summarization
>>> Hi everyone,
>>> I'm working on an analysis of illumina microarrays and am using the
>>> limma package. But looking at the code provided with the vignette there
>>> doesn't seem to be a summarization step (where multiple probes for the
>>> same gene are combined)
>>> Am I missing something here or would I need to use a different package
>>> to summarize (ie beadarray or lumi)?
>>> Thanks,
>>> -Kripa
>> ______________________________________________________________________
>> The information in this email is confidential and intended solely for the addressee.
>> You must not disclose, forward, print or use it without the permission of the sender.
>> ______________________________________________________________________

The information in this email is confidential and intend...{{dropped:4}}

More information about the Bioconductor mailing list