[BioC] MBNI custom CDF packages

Henrik Bengtsson hb at stat.berkeley.edu
Tue May 27 16:43:57 CEST 2008


Hi Manhong,

thanks for doing this big job.  In you previous releases you have
generated ASCII CDF files.  Thinking beyond R, would you mind
releasing binary CDF files instead?  They are much faster and memory
efficient to read (e.g. using Affymetrix Fusion SDK via R package
affxparser)?  A binary CDF is also much smaller (25-50% of the size).
Given an ASCII CDF, it is easy to convert it to binary (v4/XDA) using
convertCdf() in affxparser, e.g.

> library(affxparser)
> convertCdf("ASCII/AtGenome_At_REFSEQ.cdf", "AtGenome_At_REFSEQ.cdf")
> hdr <- readCdfHeader("AtGenome_At_REFSEQ.cdf")
> str(hdr)
List of 12
 $ ncols      : int 534
 $ nrows      : int 534
 $ nunits     : int 8990
 $ nqcunits   : int 13
 $ refseq     : chr ""
 $ chiptype   : chr "AtGenome_At_REFSEQ"
 $ filename   : chr "./AtGenome_At_REFSEQ.cdf"
 $ rows       : int 534
 $ cols       : int 534
 $ probesets  : int 8990
 $ qcprobesets: int 13
 $ reference  : chr ""

> file.info("ASCII/AtGenome_At_REFSEQ.cdf")$size
[1] 23214131
> file.info("AtGenome_At_REFSEQ.cdf")$size
[1] 5632960

Cheers

Henrik

On Tue, May 27, 2008 at 6:36 AM, Manhong Dai <daimh at umich.edu> wrote:
> Hi Burak,
>
>
>        Thanks a lot for your interest in our custom CDF.
>
>
>        I am working hard on version 11, because this time we planed to support
> almost all Affymetrix chips and some kinds of one-mismatch probe
> sequence alignment. The data I am processing is at Tera-byte level.
>
>
>        I expect they would be ready in a couple of weeks. If you would like to
> sign-up at http://arrayanalysis.mbni.med.umich.edu/ or yahoo mail list
> customcdf, you will get notification as soon as we release them.
>
>
> Best,
> Manhong
>
> MBNI
> University of Michigan
>
>>Hi
>>In the current BioC release (2.2), the custom CDF packages are no
>>longer hosted. I was wondering whether this will be the case in the
>>following releases or can we expect to see these packages again?
>>Thanks
>>-burak
>
> _______________________________________________
> 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