memory error: tophat-fusion-post

[angerusso]

I have ~400 samples where I ran tophat2 fusion. Now I am trying to run tophat-fusion-post which gives the following error:

Traceback (most recent call last):
File "/x86_64/tophat/v2.0.8/tophat-fusion-post", line 2627, in <module>
sys.exit(main())
File "/x86_64/tophat/v2.0.8/tophat-fusion-post", line 2604, in main
read_dist(params)
File "/x86_64/tophat/v2.0.8/tophat-fusion-post", line 1196, in read_dist
parallel_work(pids, work)
File "/x86_64/tophat/v2.0.8/tophat-fusion-post", line 731, in parallel_work
child_id = os.fork()
OSError: [Errno 12] Cannot allocate memory
Sat Jun 8 23:50:10 EDT 2013

Can anyone help me this error? Thanks so much.

[rnastar]

I am also have a similar problem. I am wondering if I am not specifying enough memory per parallel job on our cluster (currently 5G)? Any help would be appreciated here.

[2013-12-18 03:44:13] Mapping right_kept_reads.m2g_um_seg1 to genome genome with Bowtie2 (1/2)
[2013-12-18 03:52:29] Mapping right_kept_reads.m2g_um_seg2 to genome genome with Bowtie2 (2/2)
[2013-12-18 04:05:24] Searching for junctions via segment mapping
[2013-12-18 05:27:58] Retrieving sequences for splices
[2013-12-18 05:37:53] Indexing splices
Traceback (most recent call last):
  File "./tophat", line 4072, in <module>
    sys.exit(main())
  File "./tophat", line 4038, in main
    user_supplied_deletions)
  File "./tophat", line 3692, in spliced_alignment
    if reads == None or not nonzeroFile(reads):
  File "./tophat", line 1155, in nonzeroFile
    samtools_view = subprocess.Popen(samtools_view_cmd, stdout=subprocess.PIPE)
  File "/usr/local/lib/python2.7/subprocess.py", line 672, in __init__
    errread, errwrite)
  File "/usr/local/lib/python2.7/subprocess.py", line 1111, in _execute_child
    self.pid = os.fork()
OSError: [Errno 12] Cannot allocate memory

[dfornika]

I've just run an analysis on ~30 libraries and got a similar error. I was running it on a multi-user machine with ~60 GB of RAM.

Traceback (most recent call last):
  File "/home/dfornika/bin/tophat-2.0.12/tophat-fusion-post", line 2644, in <module>
    sys.exit(main())
  File "/home/dfornika/bin/tophat-2.0.12/tophat-fusion-post", line 2621, in main
    read_dist(params)
  File "/home/dfornika/bin/tophat-2.0.12/tophat-fusion-post", line 1206, in read_dist
    parallel_work(pids, work)
  File "/home/dfornika/bin/tophat-2.0.12/tophat-fusion-post", line 737, in parallel_work
    child_id = os.fork()
OSError: [Errno 12] Cannot allocate memory

I did monitor the processes somewhat using 'top' and they seemed to be using ~20% of available RAM. I'm not sure what the actual limitation is that we're running into.

It seems like tophat-fusion-post is designed to be run on a large(ish) group of samples at once. Is there any sensible way to break a large multi-sample analysis down into smaller groups? Is that likely to solve this memory issue?