Dear all,
I am trying to run RepeatModeler with multiple processors with the -pa option.
I downloaded RepeatModeler (last version available here) and installed it properly (it runs perfectly but for a reaaaaaaaaaaaally long time).
The documentation here and here says that I should be able to do so: am I missing something?
My command is RepeatModeler -pa 3 -database mydatabase and gives Unknown option: pa
Many thanks!
EDIT: When I run RepeatModeler -h it seems that this option is not supported.
NAME
RepeatModeler - Model repetitive DNA
SYNOPSIS
RepeatModeler [-options] -database <XDF Database>
DESCRIPTION
The options are:
-h(elp)
Detailed help
-database
The prefix name of a XDF formatted sequence database containing the
genomic sequence to use when building repeat models. The database
may be created with the WUBlast "xdformat" utility or with the
RepeatModeler wrapper script "BuildXDFDatabase".
-engine <abblast|wublast|ncbi>
The name of the search engine we are using. I.e abblast/wublast or
ncbi (rmblast version).
SEE ALSO
RepeatMasker, WUBlast
COPYRIGHT
Copyright 2005-2010 Institute for Systems Biology
AUTHOR
Robert Hubley <rhubley@systemsbiology.org>
Arian Smit <asmit@systemsbiology.org>
•
written
by
tlorin
3 answers
You were indeed right with your assumption. Op did not use the latest version. When using version 1.0.8, the following is shown:
-pa #
Specify the number of shared-memory processors available to this
program. RepeatModeler will use the processors to run BLAST searches
in parallel. i.e on a machine with 10 cores one might use 1 core for
the script and 9 cores for the BLAST searches by running with "-pa
9".
...
Copyright 2005-2014 Institute for Systems Biology
According to the documentation, only blast searches are parallelized, but not helpers like edgeredef, elemrdef, etc.
It seems like the problem is eleredef which does not support parallelism. Blast does use multiple processes, but it seems to be eleredef that takes so long. Round after round it takes longer and I have never seen it use more than 50% CPU.
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
123 md 20 0 3118032 2.970g 612 D 5.0 0.1 2877:59 eleredef
The helper scripts do not support parallelism, that means that -pa will only affect blast, but will have no effect on edgeredef, elemredef, etc...
It would be a very nice feature to parallelize edgeredef and friends, e.g. by using GNU parallel, but to accomplish that one would have to understand what they are doing, the man pages don't provide much information, so we'd have to take a look into the code.
edgeredef
usage: fam_def seq_list start
where seq_list is the list of sequence names, start is the index of the element to start defining families. The latter one is optional.
eleredef
usage: redef seq_list start clan_ct
where seq_list is the list of sequence names, start is the index of the element to start redefining, and clan_ct is the number to start counting the number of clans. The latter two are optional.
That is all I know about what these programs are doing. They have a seq_list parameter that could be worth a try to split into smaller packages, however I suspect that that would affect the result.
dear i had the same problem .. bt later figured it out ... you havent used the engine option which is mandatory ... 1st of all I also did the same mistake later i was able to run the program when i specified the engine
Log in to answer this question.
Hi!
I am struggling the same problem, any clue of how to solve it?
Thanks in advance,
Cristina