[BioC] ExpressionSet or MAList

[Wolfgang Huber]

Hi Dan & Gordon,

the NChannelSet class is appropriate for two-colour data and is 
supported by the Biobase package. There is an example on its use e.g. in 
the vignette of the CCl4 package. It is also used in vsn, and a more 
extensive description of workflows using it will be in the new 
Bioconductor-book that comes out in July (I know that this is not of 
much use right now).

You can also store the feature annotation in an ExpressionSet or 
NChannelSet (see the featureData slot). But this is only advantageous if 
the annotation is different for each dataset (often the case for 
home-made arrays); having it in a separate annotation package (or at 
least, annotatedDataframe) is more advantageous if you are thinking of 
processing experiments that share the same array platform (think e.g. of 
updating the probe sequence - genome mapping).

I would also recommend to take care to keep the original data files and 
the scripts that produce your R objects around; although I am not aware 
of any imminent changes, I don't think there is a 100% guarantee that 
all future implementations (e.g. many years from now) of R's 
serialisation, S4 and ExpressionSet/NChannelSet will be fully and 
painlessly backwards compatible.

Finally, note the "convert" package that provides some conversion 
methods between classes.

	Best wishes
	Wolfgang

Gordon K Smyth a écrit 06/05/2008 08:32:
> Dear Dan,
> 
> It does rather depend on the microarray platform you're using.  If 
> you're generating single channel data, such as from Affymetrix or 
> Illumina arrays, you certainly shouldn't be using MAList.  On the other 
> hand, if your data is two-colour, I wouldn't myself find ExpressionSet 
> satisfactory because it doesn't "know" anything about the structure of 
> two-colour data. Rather than ExpressionSet itself, the real alternative 
> to MAList would be a data class inheriting from ExpressionSet with 
> two-colour features added. There may be such a class already on 
> Bioconductor, but I'm not familiar with any.
> 
> Best wishes
> Gordon
> 
>> Date: Wed, 30 Apr 2008 17:26:15 +0100
>> From: Daniel Brewer <daniel.brewer at icr.ac.uk>
>> Subject: [BioC] ExpressionSet or MAList
>> To: bioconductor at stat.math.ethz.ch
>>
>> Hi,
>>
>> I am starting to think about grouping a series of microarray datasets
>> into bioconductor objects so that I can quickly look to see how a gene
>> behaves in each dataset.  The two main options seem to be to use
>> ExpressionSet or Limma's MAList.  Has anyone got an opinion on which
>> would be best to use or the advantages and disadvantages of both.
>>
>> To my mind MAList stores the annotation with the dataset which I feel is
>> an advantage whereas ExpressionSet is the base implementation for many
>> libraries.
>>
>> Dan
>>
>> -- 
>> **************************************************************
>> Daniel Brewer, Ph.D.
>> Institute of Cancer Research
>> Email: daniel.brewer at icr.ac.uk
>> **************************************************************
> 
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: 
> http://news.gmane.org/gmane.science.biology.informatics.conductor

-- 
Best wishes
  Wolfgang

------------------------------------------------------------------
Wolfgang Huber  EBI/EMBL  Cambridge UK  http://www.ebi.ac.uk/huber