[Bio] / FigKernelScripts / validate_genome_gff.pl Repository:
ViewVC logotype

View of /FigKernelScripts/validate_genome_gff.pl

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1.6 - (download) (as text) (annotate)
Mon Dec 5 18:56:38 2005 UTC (14 years, 2 months 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, myrast_rel40, 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, caBIG-05Apr06-00, 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, caBIG-13Feb06-00, rast_rel_2009_03_26, mgrast_dev_10262011, rast_rel_2008_11_24, rast_rel_2008_08_07, HEAD
Changes since 1.5: +17 -0 lines
Add license words.

#
# Copyright (c) 2003-2006 University of Chicago and Fellowship
# for Interpretations of Genomes. All Rights Reserved.
#
# This file is part of the SEED Toolkit.
# 
# The SEED Toolkit is free software. You can redistribute
# it and/or modify it under the terms of the SEED Toolkit
# Public License. 
#
# You should have received a copy of the SEED Toolkit Public License
# along with this program; if not write to the University of Chicago
# at info@ci.uchicago.edu or the Fellowship for Interpretation of
# Genomes at veronika@thefig.info or download a copy from
# http://www.theseed.org/LICENSE.TXT.
#


use strict;

use Digest::MD5;
use IPC::Open2;
use FIG;
use FigGFF;


my $usage = "validate_genome_gff [genome-gff] [< genome-gff] > modified-genome-gff\n";

my @output;

my $fig = new FIG;
my $parser = new GFFParser($fig);

my $gff;

my $input_fh;

if (@ARGV == 1)
{
    my $file = shift;
    open($input_fh, "<$file") or die "Cannot open $file: $!\n";
}
elsif (@ARGV == 0)
{
    $input_fh = \*STDIN;
}
else
{
    die $usage;
}

eval {
    $gff = $parser->parse($input_fh);
};

if ($@)
{
    print "GFF file did not parse:\n$@\n";
    exit 1;
}

warn sprintf("Parsed GFF file. Genome is %s with taxon id %s\n", $gff->genome_name, $gff->genome_id);

push(@output,
     [genus_species => $gff->genome_name],
     [taxon_id => $gff->taxon_id],
     [genome_id => $gff->genome_id],
     [taxonomy => $gff->taxonomy],
     [name => $gff->genome_name],
     [project => $gff->project]);

#
# Perform some initial validation.
#

my @errors;

$gff->taxon_id ne '' or push @errors, "Taxon ID not specified";
$gff->genome_id ne '' or push @errors, "Genome ID not specified";
$gff->genome_name ne '' or push @errors, "Genome name not specified";
$gff->project ne '' or push @errors, "Project not specified";
$gff->taxonomy ne '' or push @errors, "Taxonomy not specified";

#
# Write the contigs to a temp file.
#

my $tmp = "$FIG_Config::temp/contig.$$";
#my $tmp = "contigs.out";
open(F, ">$tmp");
for my $contig (@{$gff->contigs})
{
    print F ">$contig\n";
    print F $gff->fasta_data($contig) . "\n";
}
#print F join("\n", map { ">$_\n" . $gff->fasta_data($_) } @{$gff->contigs});
close(F);

#
# Compute checksum using index_contig_files.
#

my $pid = open2(my $index, my $flist, "$FIG_Config::bin/index_contig_files");
$pid or die "Could not run $FIG_Config::bin/index_contig_files: $!\n";

print $flist join("\t", $gff->genome_id, 1, $tmp), "\n";
close($flist);

my %contig_md5;

while (<$index>)
{
    chomp;
    my @f = split(/\t/);

    if (@f == 5)
    {
	my($id, $contig, $size, $cksum, $md5sum) = @f;

	$contig_md5{$contig} = [$size, $md5sum];
    }
}
close($index);

unlink $tmp;

my $dig = new Digest::MD5;

for my $contig (sort keys %contig_md5)
{
    my($size, $md5sum) = @{$contig_md5{$contig}};
    my $l = "$contig\t$size\t$md5sum\n";
    $dig->add($l);
}

my $md5sum = $dig->hexdigest();

my $file_md5sum = $gff->genome_checksum();

if ($file_md5sum =~ /^[0-9A-Fa-f]{32}$/)
{
    if ($file_md5sum eq $md5sum)
    {
	push(@output, [genome_md5 =>  $file_md5sum]);
    }
    else
    {
	push(@errors, "Checksum provided does not match computed checksum");
    }
}
else
{
    push @errors, "Checksum not specified (computed as $md5sum)";
}

if (@errors)
{
    print join("\n", @errors), "\n";
    exit(1);
}

for my $elt (@output)
{
    print join("\t", @$elt), "\n";
}


MCS Webmaster
ViewVC Help
Powered by ViewVC 1.0.3