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

View of /FigKernelScripts/make_sp_uni_files.pl

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1.4 - (download) (as text) (annotate)
Thu Jun 14 18:32:55 2007 UTC (12 years, 8 months ago) by wilke
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, 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
Changes since 1.3: +6 -5 lines
changed check for input and output directory
the directories may exists, so don't die

# -*- perl -*-
#
# 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 Cwd;
use File::Path;

use FIG;

$0 =~ m/([^\/]+)$/;
$usage = "$1 [-noget] InDir OutDir";

$noget = "";
if ($ARGV[0] && ($ARGV[0] =~ m/-noget/)) { $noget = shift; }

(($indir  = shift @ARGV) && ($outdir = shift @ARGV))
    || die "\n\tusage: $usage\n\n";

if ($noget) {
    (-d  "$indir/SwissProt") || die "Input directory $indir/SwissProt does not exist";
} else {
    (!-d "$indir/SwissProt") || die "Input directory $indir/SwissProt already exists";
    mkpath("$indir/SwissProt",  1, 0777) || die "Could not create $outdir/SwissProt";
}

if (!-d "$outdir/SwissProt") {
    mkpath("$outdir/SwissProt", 1, 0777) || die "Could not create $outdir/SwissProt";
}
if (!-d "$outdir/UniProt") {
    mkpath("$outdir/UniProt",   1, 0777) || die "Could not create $outdir/UniProt";
}

$base = getcwd();


#+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
# get data from ftp://us.expasy.org/databases/swiss-prot/release/
#-----------------------------------------------------------------------
$sp_url = "ftp://us.expasy.org/databases/swiss-prot/release/";
unless ($noget)
{
    chdir("$indir/SwissProt") || die "Could not chdir to $indir/SwissProt/, return-code $!";
    if (($test = `which wget`) && ($test !~ m/^which: no wget in/))
    {
	&FIG::run("wget --passive-ftp $sp_url/README");
	&FIG::run("wget --passive-ftp $sp_url/reldate.txt");
	&FIG::run("wget --passive-ftp $sp_url/uniprot_sprot.dat.gz");
	&FIG::run("wget --passive-ftp $sp_url/uniprot_trembl.dat.gz");
    }
    elsif (($test = `which curl`) && ($test !~ m/^which: no curl in/))
    {
	&FIG::run("curl --ftp-pasv -f -O $sp_url/README");
	&FIG::run("curl --ftp-pasv -f -O $sp_url/reldate.txt");
	&FIG::run("curl --ftp-pasv -f -O $sp_url/uniprot_sprot.dat.gz");
	&FIG::run("curl --ftp-pasv -f -O $sp_url/uniprot_trembl.dat.gz");
    }
    else
    {
	die "Sorry, could not find either \`wget\` or \`curl\` in path $ENV{PATH}";
    }
    chdir($base)  || die "Could not chdir to $base: $!";
}

if (-s "$indir/SwissProt/uniprot_sprot.dat.gz")
{
    print STDERR "Processing $indir/SwissProt/uniprot_sprot.dat.gz\n";
    &FIG::run("gunzip -c $indir/SwissProt/uniprot_sprot.dat.gz | sp2fasta -l sp        > $outdir/SwissProt/fasta");
    &FIG::run("gunzip -c $indir/SwissProt/uniprot_sprot.dat.gz | make_sp_org_table sp  > $outdir/SwissProt/org.table");
    &FIG::run("gunzip -c $indir/SwissProt/uniprot_sprot.dat.gz | make_sp_desc_with_org > $outdir/SwissProt/sp.desc");
}
elsif (-s "$indir/SwissProt/uniprot_sprot.dat")
{
    print STDERR "Processing $indir/SwissProt/uniprot_sprot.dat\n";
    &FIG::run("sp2fasta -l sp        < $indir/SwissProt/uniprot_sprot.dat > $outdir/SwissProt/fasta");
    &FIG::run("make_sp_org_table sp  < $indir/SwissProt/uniprot_sprot.dat > $outdir/SwissProt/org.table");
    &FIG::run("make_sp_desc_with_org < $indir/SwissProt/uniprot_sprot.dat > $outdir/SwissProt/sp.desc");
}
else
{
    die "Could not find either $indir/SwissProt/uniprot_sprot.dat.gz or $indir/SwissProt/uniprot_sprot.dat";
}

if (-s "$indir/SwissProt/uniprot_trembl.dat.gz")
{
    print STDERR "Processing $indir/SwissProt/uniprot_trembl.dat.gz\n";
    &FIG::run("gunzip -c $indir/SwissProt/uniprot_trembl.dat.gz | sp2fasta -l tr           >> $outdir/SwissProt/fasta");
    &FIG::run("gunzip -c $indir/SwissProt/uniprot_trembl.dat.gz | make_sp_org_table tr     >> $outdir/SwissProt/org.table");
    &FIG::run("gunzip -c $indir/SwissProt/uniprot_trembl.dat.gz | make_sp_desc_with_org tr >> $outdir/SwissProt/sp.desc");
}
elsif (-s "$indir/SwissProt/uniprot_trembl.dat")
{
    print STDERR "Processing $indir/SwissProt/uniprot_trembl.dat\n";
    &FIG::run("sp2fasta -l tr           < $indir/SwissProt/uniprot_trembl.dat >> $outdir/SwissProt/fasta");
    &FIG::run("make_sp_org_table tr     < $indir/SwissProt/uniprot_trembl.dat >> $outdir/SwissProt/org.table");
    &FIG::run("make_sp_desc_with_org tr < $indir/SwissProt/uniprot_trembl.dat >> $outdir/SwissProt/sp.desc");
}
else
{
    die "Could not find either $indir/SwissProt/uniprot_trembl.dat.gz or $indir/SwissProt/uniprot_trembl.dat";
}

# NOTE: trembl_new.dat.gz no longer appears to exist as a separate category...

print STDERR "Extracting SP funcs and aliases\n";
&FIG::run("make_sp_func_syn_dicts $outdir/SwissProt < $outdir/SwissProt/sp.desc");


#+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
print STDERR "Creating corresponding UniProt files:\n";
#-----------------------------------------------------------------------
print STDERR "$outdir/UniProt/assigned_functions ...\n";
open(SWISS_IN, "<$outdir/SwissProt/assigned_functions") 
    || die "Could not read-open $outdir/SwissProt/assigned_functions";
open(UNI_OUT,  ">$outdir/UniProt/assigned_functions") 
    || die "Could not write-open $outdir/UniProt/assigned_functions";
while (defined($entry = <SWISS_IN>))
{
    $entry =~ s/^(sp|tr)/uni/;
    print UNI_OUT $entry;
}
close(UNI_OUT)  || die "Could not close $outdir/UniProt/assigned_functions";
close(SWISS_IN) || die "Could not close $outdir/SwissProt/assigned_functions";

print STDERR "$outdir/UniProt/org.table ...\n";
open(SWISS_IN, "<$outdir/SwissProt/org.table") 
    || die "Could not read-open $outdir/SwissProt/org.table";
open(UNI_OUT,  ">$outdir/UniProt/org.table") 
    || die "Could not write-open $outdir/UniProt/org.table";
while (defined($entry = <SWISS_IN>))
{
    $entry =~ s/^(sp|tr)/uni/;
    print UNI_OUT $entry;
}
close(UNI_OUT)  || die "Could not close $outdir/UniProt/org.table";
close(SWISS_IN) || die "Could not close $outdir/SwissProt/org.table";

print STDERR "$outdir/UniProt/fasta ...\n";
open(SWISS_IN, "<$outdir/SwissProt/fasta") 
    || die "Could not read-open $outdir/SwissProt/fasta";
open(UNI_OUT,  ">$outdir/UniProt/fasta") 
    || die "Could not write-open $outdir/UniProt/fasta";
while (($id, $seqP) = &FIG::read_fasta_record(\*SWISS_IN))
{
    $id =~ s/^(sp|tr)/uni/;
    &FIG::display_id_and_seq($id, $seqP, \*UNI_OUT);
}
close(UNI_OUT)  || die "Could not close $outdir/UniProt/fasta";
close(SWISS_IN) || die "Could not close $outdir/SwissProt/fasta";

exit(0);

MCS Webmaster
ViewVC Help
Powered by ViewVC 1.0.3