[Bioc-devel] MIAPE/MIAME in eSet

[Martin Morgan]

On 10/13/2010 04:04 AM, Vincent Carey wrote:
> I don't know of any plans to generalize this, but I would tend to
> agree that eSet should not commit one to a microarray-oriented
> experiment metadata element.  I looked briefly at the PSI resources
> and see that schemata have been generated for various assay types.  So
> it's not the case that there's a simple way of comparing the PSI
> schema to what is already
> present in MIAME.
> 
> Currently we have
> 
>> getClass("MIAME")
> Class "MIAME" [package "Biobase"]
> 
> Slots:
> 
> Name:               name               lab           contact             title
> Class:         character         character         character         character
> 
> Name:           abstract               url         pubMedIds           samples
> Class:         character         character         character              list
> 
> Name:     hybridizations      normControls     preprocessing             other
> Class:              list              list              list              list
> 
> Name:  .__classVersion__
> Class:          Versions
> 
> Extends: "Versioned", "characterORMIAME"
> 
> All but "hybridizations" would seem potentially relevant to any
> high-throughput experimental paradigm.  So except for some naming
> inelegance, I would propose that MIAPE class design proceed by
> extending MIAME.  Your eSet extensions can load the experimentData
> slot with MIAPE-CC, -MS and so forth, as you like
> (http://www.psidev.info/index.php?q=node/91), but we would leave the
> class of experimentData as MIAME to avoid ramifications of changing a
> central class definition.
> 
> This is my humble opinion only, and more discussion is clearly
> warranted.  I would conclude by saying that IMHO changing the eSet

Certainly extending MIAME would be backward compatible; from brief
reflection it seems like a virtual class MIA at the base of MIAME /
MIAPE / MIA*E could also be inserted without disrupting current
serialized instances.

As Laurent mentions the euro-dev meeting would be a good forum to
formalize plans, but we on our end will give this some attention over
the next couple of weeks.

Martin

> definition is not to be taken lightly, and functional benefits of such
> a change need to be very substantial to warrant this.
> 
> 
> On Wed, Oct 13, 2010 at 6:03 AM, Florian P. Breitwieser
> <fbreitwieser at cemm.oeaw.ac.at> wrote:
>>  Dear developers,
>>
>> we are developing a package for Bioconductor, using eSet as a container for
>> the assay data.
>>
>> It is Mass spectrometry proteomics data. In the same manner as MIAME exists
>> for Microarrays, there is a minimum information about a proteomics
>> experiment (MIAPE) defined by the Proteomics Standards initiative.
>>
>> Currently, experimentData slot in eSet is of class MIAME. Are there plans to
>> generalize this?
>> I am happy to help, and to contribute a MIAPE class definition.
>>
>> Best,
>> Florian P Breitwieser
>>
>> _______________________________________________
>> Bioc-devel at stat.math.ethz.ch mailing list
>> https://stat.ethz.ch/mailman/listinfo/bioc-devel
>>
> 
> _______________________________________________
> Bioc-devel at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/bioc-devel


-- 
Computational Biology
Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N. PO Box 19024 Seattle, WA 98109

Location: M1-B861
Telephone: 206 667-2793