[BioC] rMAT tutorial. Data file too large?

[Steve Lianoglou]

Hi,

On Tue, Apr 27, 2010 at 11:01 AM, Patrick Schorderet
<patrick.schorderet at epfl.ch> wrote:
>
> Hello everybody,
>
> I wanted to use rMAT for some of my ChIP-chip analysis and tried to follow
> the tutorial (http://wiki.rglab.org/index.php?title=Public:RMAT). I have
> updated everything (R version 2.11.0 and rMAT version 2.4.0) and have
> downloaded the tutorial data file (EM). I have managed to do everything up
> to the rbind part. I have the same summary as in the tutorial (summary(ER)),
> but as soon as I try to normalize the data, my R crashes. I am wondering if
> any of you have experienced similar outcomes? Could the amount of processing
> be too large? Has any of you been able to successfully follow the tutorial?

I don't know what you mean by "crash", but:

I tried to run it on my machine. I'm not sure if it's crashing, but
the call to `ERnorm<-NormalizeProbes(ER,
method="PairBinned",robust=TRUE)` pegs a CPU @ 100% for a long while
and doesn't return after what I think has been at least 20 mins, or so
(I've killed the process).

For the record, I'm running 64bit R on a Mac Pro with 8 GB of RAM. I
noticed that you are running 32bit R and the memory of the pegged R
process is upwards of 2GB (1.9GB to be exact). Since 32bit R is stuck
with dealing with 2GB of RAM at a stretch, perhaps this might be one
of the problems you are facing.

So ... if you do have a 64bit system with enough memory, perhaps you
can try running R in 64bit mode (use R64.app if you want to use the
GUI) to see if the normalization step ever finishes.

The versions of our packages should be the same since I just updated
to R 2.11 (release) yesterday and updated all packages along with it.

-steve

-- 
Steve Lianoglou
Graduate Student: Computational Systems Biology
 | Memorial Sloan-Kettering Cancer Center
 | Weill Medical College of Cornell University
Contact Info: http://cbio.mskcc.org/~lianos/contact