Re: [Mged-mage] Comprehending the BioDataCube

[Helen P. <par...@eb...>]

Valerie

we discourage use of tuples the files are huge and v. slow when loading 
- we have never successfully loaded data of this typ. Most data is coded 
as data external - we prefer this as it's easier to detect a problem 
data file e.g. with a simple wc when we have a problem.


The tab delimited files  3 dimensions are represented by the 
QuantitationType (QTD), DesignElement (DED) and BioAssayDimensions 
(BAD), these are expressed in the MAGE-ML and provided the dimensions 
needed to understand these tab delimited files.

The QTD which typically governs the no of columns is made of QTs that 
are determined by the sw that generated the matrix. All the raw data QT 
that we are aware of from common sw types are detailed on this 
spreadsheet and these are the recommended best practice coding for these 
SW types.


http://www.ebi.ac.uk/~ele/ext/submitter.html#qt

There is also some other information there that may be useful on the 
data cube. Note also that early cel files were transformed to tab 
delimited or space delimited data. We no longer do this, and use the 
native format cel files instead, though the dimensions described above 
are still provided. This allows us to check that the cel file supplied 
is corresponds to what was intended.

Please ask if we can provide more information.

best regards

Helen

Valerie Wagner wrote:
> Folks, can anyone help me understand the meaning of the data in 
> BioDataValues?
>
> What I understand:
>
> * BioDataCube is a 3-dimensional representation of the BDQ data, where 
> BDQ = BioAssays, DesignElements, QuantiationTypes.  However, that's not 
> a very specific explanation of what the data mean.
>
> * MAGE-ML differs from the MAGE OM in how it represents this data, using 
> instead DataInternal and DataExternal.
>
> What I don't understand:
>
> * Is the format of the external data files subject to a MAGE standard? 
> All the examples I've seen so far are "tab delimited", but it doesn't 
> look to me like the data files all have the same format (differing 
> number of columns, some have all numeric data, some mixed data, etc.). 
> How does one tell what one is looking at?
>
> * Similarly, all the examples I've found so far use the external data 
> method.  Does anyone have an example of anything using BioDataTuples 
> instead of the cube or using DataInternal instead of external?
>
> Thanks again!
> Valerie
>
>
>
> -------------------------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job easier
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> Mged-mage mailing list
> Mge...@li...
> https://lists.sourceforge.net/lists/listinfo/mged-mage
>   

-- 
Helen Parkinson, PhD
Curation Coordinator
Microarray Informatics Team, 
EBI

and

Seconded Scientific Programme Manager
NCRI Cancer Informatics Initiative
www.cancerinformatics.org.uk

Tel:
EBI 01223 494672
Skype: helen.parkinson.ebi