[Bio] / FortyEight / rp_gen_teach_sims.pl Repository:
ViewVC logotype

View of /FortyEight/rp_gen_teach_sims.pl

Parent Directory Parent Directory | Revision Log Revision Log

Revision 1.1 - (download) (as text) (annotate)
Fri Oct 5 17:25:53 2007 UTC (12 years, 1 month ago) by olson
Branch: MAIN
CVS Tags: mgrast_dev_08112011, rast_rel_2009_05_18, mgrast_dev_08022011, rast_rel_2014_0912, rast_rel_2008_06_18, rast_rel_2008_06_16, mgrast_dev_05262011, rast_rel_2008_12_18, mgrast_dev_04082011, rast_rel_2008_07_21, rast_rel_2010_0928, rast_2008_0924, mgrast_version_3_2, mgrast_dev_12152011, rast_rel_2008_04_23, mgrast_dev_06072011, rast_rel_2008_09_30, rast_rel_2009_0925, rast_rel_2010_0526, rast_rel_2014_0729, mgrast_dev_02212011, rast_rel_2010_1206, mgrast_release_3_0, mgrast_dev_03252011, rast_rel_2010_0118, mgrast_rel_2008_0924, mgrast_rel_2008_1110_v2, rast_rel_2009_02_05, rast_rel_2011_0119, mgrast_rel_2008_0625, mgrast_release_3_0_4, mgrast_release_3_0_2, mgrast_release_3_0_3, mgrast_release_3_0_1, mgrast_dev_03312011, mgrast_release_3_1_2, mgrast_release_3_1_1, mgrast_release_3_1_0, mgrast_dev_04132011, rast_rel_2008_10_09, mgrast_dev_04012011, rast_release_2008_09_29, mgrast_rel_2008_0806, mgrast_rel_2008_0923, mgrast_rel_2008_0919, rast_rel_2009_07_09, rast_rel_2010_0827, mgrast_rel_2008_1110, myrast_33, rast_rel_2011_0928, rast_rel_2008_09_29, mgrast_rel_2008_0917, rast_rel_2008_10_29, mgrast_dev_04052011, mgrast_dev_02222011, rast_rel_2009_03_26, mgrast_dev_10262011, rast_rel_2008_11_24, rast_rel_2008_08_07, HEAD
Updates to the Teach stuff. Small mod to build_nr to support condtional setting of sort buffer size.

# Given the computed glimmer and critica calls from earlier runs of rp_glimmer
# and rp_critica, use build_nr to compute the mappings between these calls and
# the peg calls in the genome, extract the sims that we can from the
# precomputed sims, and compute the remaining sims as needed.

use Data::Dumper;
use Carp;
use strict;
use Job48;
use FIGV;
use FIG;
use FIG_Config;
use File::Basename;
use GenomeMeta;
use Sim;

@ARGV == 1 or die "Usage: $0 job-dir\n";

my $jobdir = shift;

-d $jobdir or die "$0: job dir $jobdir does not exist\n";

my $job = new Job48($jobdir);
$job or die "cannot create job for $jobdir";

my $hostname = `hostname`;
chomp $hostname;

my $genome = $job->genome_id();
my $meta = $job->meta();

$meta->add_log_entry($0, "start teach-sims processing on $hostname in $jobdir");

my $work = "$jobdir/teach_sims_work";
my $orgdir = "$jobdir/rp/$genome";

chdir $work or &fatal("chdir $work failed: $!");

open(SRCS, ">$work/nr.sources") or &fatal("cannot write $work/nr.sources: $!");

for my $type (qw(peg glimmer critica))
    print SRCS "$orgdir/Features/$type/fasta\n";

     "$work/nr.sources", "/dev/null", "/dev/null",
     "$work/teach.nr", "$work/teach.ps");

# Read the generated peg.synonyms. We are going to build a list
# @sims_to_compute of xxx ids that do not have peg ids.
# Also compute the mapping %peg_to_teach of pegid => [id]
# for the ids that do map to pegs.

my @sims_to_compute;
my %peg_to_teach;
my %xxx_to_teach;

open(PULL, "| $FIG_Config::bin/pull_fasta_entries $work/teach.nr > $work/sims.nr") or
    &fatal("cannot open pull pipeline: $!");

open(PS, "<$work/teach.ps") or &fatal("Cannot open $work/teach.ps: $!");

while (<PS>)
    if (/^([^,]+),(\d+)\t(.*)/)
	my $ps = $1;
	my $ps_len = $2;
	my @pairs = map { [ split(/,/, $_) ] } split(/;/, $3);

	my @pegs = grep { $_->[0] =~ /^fig\|\d+\.\d+\.peg/ } @pairs;
	my @rest = grep { $_->[0] !~ /^fig\|\d+\.\d+\.peg/ } @pairs;
#	print Dumper(\@pairs, \@pegs, \@rest);
	if (@pegs)
	    my $rep = $pegs[0]->[0];
	    $peg_to_teach{$rep} = [@rest];
	    push(@sims_to_compute, $ps);
	    $xxx_to_teach{$ps} = [@rest];
	    print PULL "$ps\n";
print Dumper(\@sims_to_compute, \%xxx_to_teach, \%peg_to_teach);
if (!close(PULL))
    &fatal("error closing pull pipeline; \$!=$! \$?=$?");

$meta->add_log_entry($0, "finish critica computation on $jobdir");
$meta->set_metadata("status.critica", "complete");

sub run
    my(@cmd) = @_;
    print "Run @cmd\n";
    my $rc = system(@cmd);
    $rc == 0 or &fatal("Cmd failed with rc=$rc: @cmd");

sub fatal
    my($msg) = @_;

    if ($meta)
	$meta->add_log_entry($0, ['fatal error', $msg]);
	$meta->set_metadata("status.bbhs", "error");

    croak "$0: $msg";

MCS Webmaster
ViewVC Help
Powered by ViewVC 1.0.3