gcrma() and get() -- oops, 2nd try [sorry for 1st bogus send]
rfinney5 at yahoo.com
Wed Mar 9 21:27:52 CET 2005
Thanks, Zhijin, I have, through a backdoor,
created the probepackage for a custom CDF.
I do have a second question:
gcrma() uses the get() function/method.
Is there a way to force get() to look locally
and not in the "repository" out on the internet?
I can't quite figure out the invocation of this
bit of magic.
[ thanks folks and sorry for the previous
mis-communication. i hit the wrong button. ]
--- Zhijin Wu <zwu at jhsph.edu> wrote:
> gcrma uses probe sequence information in estimating
> non-specific binding
> so you need the probepackge. You can create the
> probepackge if it's not
> available on bioconductor. The vignette for
> 'matchprobes" package has
> description of the function "makeProbePackage".
> Here's the link
> On Tue, 8 Mar 2005, Richard Finney wrote:
> > I am having a little trouble torturing gcrma()
> > into accepting a custom CDF file created locally.
> > (Local in this case is R Ver 2.0.1 on a reasonably
> > recent linux box).
> > I can create and convince R/Bioc to accept
> > a custom CDF file in some cases. It works
> > fine with rma() and mas5() . But ... gcrma
> > uses some "probepackage" thingamajig and goes
> > looking out on the internet for some custom
> > name for the custom cdf library I created and
> > of course does not find it.
> > What function/method do I need to run to create
> > a custom "probe information" file for my custom
> > CDF file how do I make gcrma() use this new
> > information by looking locally for it and not
> > out on the internet?
> > Help. pleez.
> > -rich
> > _______________________________________________
> > Bioconductor mailing list
> > Bioconductor at stat.math.ethz.ch
> > https://stat.ethz.ch/mailman/listinfo/bioconductor
More information about the Bioconductor