[rmaexpress_help] Re: Difficulty in reading files

  • From: Jun Ding <junding@xxxxxxxxx>
  • To: rmaexpress_help@xxxxxxxxxxxxx
  • Date: Fri, 01 Sep 2006 12:20:39 -0400

Hi Ben,

Thank you very much for your help and for updating the software! Your email is very helpful and I will try your first method in the coming days. I believe by doing that I should be able to process the data. I will let you know what happens.

Thank you again!

Best,
Jun
----------------------------
Jun Ding, Ph.D. student
Department of Biostatistics
University of Michigan
Ann Arbor, MI, 48105
----------------------------


Quoting Ben Bolstad <bmb@xxxxxxxxxxxxx>:

Hi Jun,

There are two separate issues going on here. The first is a problem with
the RMAExpress code which I will go about fixing immediately. Basically
what is happening is that it is detecting an error with the CEL file,
but failing to properly close a modal dialog box and reactivate the main
memus so it locks up.

The second issue here is actually the underlying cause of the problem.
Specifically, some of the arrays are HG_U95A and some are HG_U95Av2.
RMAExpress tries to ensure that all CEL files being read in are from the
same array type and in this case it is finding a mismatch. In particular
these are the array types for that set of files.


GSM60097.CEL: HG_U95Av2 GSM60098.CEL: HG_U95Av2 GSM60099.CEL: HG_U95A GSM60100.CEL: HG_U95A GSM60101.CEL: HG_U95A GSM60102.CEL: HG_U95A GSM60103.CEL: HG_U95A GSM60104.CEL: HG_U95A GSM60105.CEL: HG_U95A GSM60106.CEL: HG_U95A GSM60107.CEL: HG_U95A

Now this is really a special case because the U95A and U95Av2 differ
based upon only a relative handful of probesets. There are several ways
to handle this. The best way is to use the RMADataConv program basically
following the instructions in the user guide in the subsection "Merging
MG U74A and MG U74Av2 datasets" except substituting HG_U95A and
HG_U95Av2 in place for the mouse array names. With the requisite file of
overlap names available at:

http://bmbolstad.com/misc/mixtureCDF/HGU95Aoverlap.txt

The second way to handle this (and not necessarily recommended) is to
manually edit the part of the CEL files which describes the name of
appropriate array type so that they are all identical and just pretend
that all the arrays are really of one type only.

Hope this helps,

Ben




On Thu, 2006-08-31 at 16:36 -0400, Jun Ding wrote:
Hi RMAExpress Developers and Users,

I have difficulties in reading the .CEL files into RMAExpress. All the
.CEL (11) files are downloaded (GEO dataset GSE2737) and extracted, but
I can only read in the first two .CEL files. When I tried to read in
more (or only read one of the other files), RMAExpress just crashed. I
have tried RMAExpress release 0.4.1 and 0.4 alpha 7. I also got an
error message from Visual Studio Just-In-Time Debugger: "An unhandled
win32 exception occurred in RMAExpress.exe [3352]."

Does that mean all but 2 .CEL files are corrupted? Can anyone give me
some suggestions? Thank you very much!

Best,
Jun

----------------------------
Jun Ding, Ph.D. student
Department of Biostatistics
University of Michigan
Ann Arbor, MI, 48105
----------------------------





--
Ben Bolstad <bmb@xxxxxxxxxxxxx>
http://bmbolstad.com








Other related posts: