[BioC] Re: offer from I Neuhaus to create a BMS response; kee p in core, please

Warnes, Gregory R gregory_r_warnes at groton.pfizer.com
Tue Jul 1 10:21:54 MEST 2003


Would it not be a better mechanism to just have them insert a
'content-creator' field into thier binary format?   Several of my collegues
have championed the idea of using Affy's CEL file as a standard format for
distributing probe-level data, specificially because one could then open the
file in Affy's standard tools after doing some other type of processing.
For example, zeroing out the MM probes to investigate the effect on the Affy
algorithms*.  

-Greg 

* Last time we checked, this would cause a divide by zero error, as would
replacing with any other fixed value, presumably because of a variance
calculation returning zero.  Putting in small non-constant non-zeros worked
and produced some interesting results.  

-Greg

> -----Original Message-----
> From: Rafael A. Irizarry [mailto:ririzarr at jhsph.edu]
> Sent: Tuesday, July 01, 2003 9:14 AM
> To: bioconductor at stat.math.ethz.ch
> Subject: [BioC] Re: offer from I Neuhaus to create a BMS 
> response; keep
> in core, please
> 
> 
> On Tue, 1 Jul 2003, Robert Gentleman wrote:
> 
> <SNIP>
> > Vis a vis the phone call, it was about 1 hour long, it was slightly
> > focussed on marketing hype on their side and some substantial,
> > although peculiar concerns came out. One of the major concerns that
> > they have is that if they make their formats public then anyone can
> > create a "CEL" file of the right format and make it look very much
> > like one that they generated. They believe that this would lead to
> > substantial confusion. To some extent I agree with them and I think
> > that I would be inclined to discourage any such activity, and I hope
> > BioC as a group would too (since it really spreads confusion, when
> > opening a CEL file we want to be pretty sure where it came from and
> > what is in it).
> > 
> 
> i just rememebered, in the affy package there is a function called 
> 
> write.celfile
> 
> should we remove this?
> 
> 
> <SNIP>
> 
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor
> 


LEGAL NOTICE\ Unless expressly stated otherwise, this messag...{{dropped}}



More information about the Bioconductor mailing list