[BioC] illumina --> limma?

Mark Dunning md392 at cam.ac.uk
Fri Aug 25 17:04:47 CEST 2006


So just to clarify, here is the situation regarding BeadArray raw data.

If you have made the appropriate tweaks to the BeadScan software then  
you should get a text file that look like this (for each array/strip);

Code    Grn    GrnX    GrnY
50008    796    536.5297    3167.486
50008    495    1574.66    12360.35
50008    706    2120.334    14611.3
50008    519    1493.311    17420.35
50008    938    1549.49    11478.03
50008    612    1989.498    4735.532
50008    744    1073.537    15349.49
..........

*********************************************
For instructions on how to make the tweaks in BeadScan go to

http://www.damtp.cam.ac.uk/user/npt22

On that page you will find an xml file to replace the existing  
Settings.xml file within the BeadScan program directory

Note that these steps can only be used before scanning new arrays. I  
don't know of any way to recover the x and y coordinates for arrays  
that have been previously scanned.

  *********************************************


As Sean notes, each bead is identified by a Code and not a TargetID  
as seen in the BeadStudio output. This is annoying, but it is  
possible to match up the two identifiers using the bead set manifest  
that Illumina distributes on their cd's when they give the arrays   
(its a csv file).


The Grn column gives the bead intensity that has been calculated  
using Illumina image processing methods which include a "sharpening"  
transformation and a local background correction. It is possible for  
both these steps to produce negative values. The beadarray package is  
able to use the TIF images and the X and Y coordinates from the text  
file to calculate bead intensities. It is possible to re-create the  
steps that Illumina use although we store the local foreground and  
background for each bead separately so that background correction is  
optional. We are also looking to implement a method of background  
estimation using morphological filtering.

We have only just seen this new text format for ourselves, so the  
code to read these files is only available in the latest version of  
beadarray in the developers section.

So in summary, with the Code, X and Y information plus the .tif  
image, you can re-create the Illumina image analysis yourself to get  
raw intensities at the probe level using the beadarray package.
Currently this takes around 1min for each array (strip) i.e. about  
10-15mins for each chip [ 3Ghz machine with 3GB ram].

Hope this helps,



Mark

On 25 Aug 2006, at 14:55, Stefano Calza wrote:

>
>
> On Fri, Aug 25, 2006 at 09:42:36AM -0400, Sean Davis wrote:
> <Sean>Mark,
> <Sean>
> <Sean>Have you made any headway with Illumina regarding getting  
> data that is
> <Sean>closer to "raw" but still useful?  We are getting the probe- 
> level data
> <Sean>dumped from our (which I think is NOT background-subtracted,  
> but correct me
> <Sean>if I am wrong), but these data do not conform to the usual  
> Illumina output
> <Sean>(different unique ids for the probes), so I really haven't  
> given much
> <Sean>thought to using them, particularly since I can't convince  
> our collaborators
> <Sean>to make hacks of their scanner config file to get the dumps  
> for all the data
> <Sean>we receive.
> <Sean>
>
>
> I guess we are getting something like that, but afaik the dumps are  
> already somehow corrected (and I get some negative values). The  
> scanner itself may do some background correction when dumping numbers.
> You may need to use the raw scanned image, but it may not be  
> possible (if not through an agreement with Illumia).
>
> I'm just trying to understand what's in this data, so correct me if  
> I'm wrong.
>
> Regards,
> Stefano
>
> -- 
> Stefano Calza, PhD
> Researcher - Biostatistician
> Sezione di Statistica Medica e Biometria
> Dipartimento di Scienze Biomediche e Biotecnologie
> Università degli Studi di Brescia - Italy
> Viale Europa, 11 25123 Brescia
> email: calza at med.unibs.it
> Phone: +390303717653
> Fax: +390303717488
>
> _______________________________________________
> 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
>



More information about the Bioconductor mailing list