Having trouble running PrepareAllPathsInputs.pl in AllPathsLG

[marrakesh]

Hi! I'm posting after having read as many threads here as I can on similar problems, but I still can't get this script to run. I'm trying to run a de novo assembly of a mammalian genome for the first time and was instructed to use Allpaths LG on my data. Unfortunately, I'm having some difficulty getting PrepareAllPathsInputs.pl to run so that I can even try to figure out Allpaths. Googling hasn't yielded any solutions.

When I run the program, I get an output that starts having problems here:

Call to new failed, memory usage before call = 22473992k.

Here is the output of top:

/bin/sh: fork: Cannot allocate memory
Aborting.

Thu Jul 31 13:28:07 2014. Abort. Stopping.

Generating a backtrace...
Bummer. I can't run ps to determine the command name.
failed: ps -p 2309 -o command
Here's what happened:
Exiting.

Thu Jul 31 13:28:07 2014. Abort. Stopping.

Generating a backtrace...
Bummer. I can't run ps to determine the command name.
failed: ps -p 2309 -o command
Here's what happened:
Exiting.

Thu Jul 31 13:28:07 2014. Abort. Stopping.

Generating a backtrace...
Bummer. I can't run ps to determine the command name.
failed: ps -p 2309 -o command
Here's what happened:
Exiting.

Those repeat for a while, and then I get this:

Generating a backtrace...
/bin/sh: /usr/bin/top: Argument list too long

Dump of stack:

0. std::basic_stringbuf<char, std::char_traits<char>, std::allocator<char> >::~b asic_stringbuf( ... ), in ??:0
1. std::basic_stringbuf<char, std::char_traits<char>, std::allocator<char> >::~b asic_stringbuf( ... ), in ??:0
(cut--this repeats all the way up to n=33)

There's a little more in here, but it either repeats previous things or doesn't appear to be an error message, and I'm cutting in the interests of brevity. Finally the error message ends:

**** 2014-07-31 13:39:41 (CG): 2 errors occurred while converting groups. Aborting.

Thu Jul 31 13:39:41 2014 - /opt/apps/allpathslg/44837/bin/PrepareAllPathsInputs. pl
A shell command returned exit value 29.
Exiting Perl.

This is my script:

PrepareAllPathsInputs.pl DATA_DIR=/scratch/02985/emg2497/mouse_genome_project/al l_paths_analysis/S.teguina/data/ PICARD_TOOLS_DIR=/opt/picard/bin/ PLOIDY=2 PHRED_64=0

I can also provide my all_groups and all_libs .csv files if that would be helpful. I've made sure that all fields are filled and that I have a space after the final comma on all of the entries.

Strangely, the Perl script seems to have worked fine on my Pac-Bio reads, but not my Illumina fragment and mate-pair libraries. I have no idea why this would be the case.

Does anyone have any insight at all? I am at my wits' end.

[GenoMax]

Looks like the process is running out of memory (seems to reach 21G usage based on message above). How much RAM do you have available on this machine?

[marrakesh]

That's a bit tricky, because I'm not running it on a computer per se--I'm running on a server allocation that's remote. I had thought that the server was allocating unlimited RAM to the batch job I was submitting--in part because of the quote below from the output file--but maybe I'm wrong about that.

TACC: Setting memory limits for job 2097802 to unlimited KB

The 23 GB number you mention pops up earlier, so I need to try and figure out what's up with the server allocation I'm working on. Thanks for the tip!