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

View of /FigKernelScripts/synchronize_annotations_and_assignments.pl

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1.2 - (download) (as text) (annotate)
Mon Dec 5 18:56:38 2005 UTC (13 years, 11 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 1.1: +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.
#


#
# Create an annotation for every function assignment that does not currently have one.
#
# Load the assigned_functions file for a genome into memory.
#
# Walk the annotations file, loading the most recent master annotation for each peg.
#
# Then walk the two lists, comparing and writing out annotations for the ones that are
# missing.
#

use File::stat;
use Data::Dumper;
use FileHandle;
use strict;
use FIG;
my $fig = new FIG;

my %gcount;

if (@ARGV < 3)
{
    die "Usage: $0 anno_file assign_file genome [genome ..]\n";
}

my $anno_file = shift;
my $assign_file = shift;

my @genomes = @ARGV;


if (@genomes == 1 and $genomes[0] eq '-')
{
    @genomes = ();
    while (<STDIN>)
    {
	chomp;
	if (/^\s*(\d+\.\d+)/)
	{
	    push(@genomes, $1);
	}
    }
}

for my $genome (@genomes)
{
    if ($genome !~ /^\d+\.\d+$/)
    {
	die "Invalid genome $genome\n";
    }
}

{
    my $anno_fh = new FileHandle(">$anno_file");
    $anno_fh or die "Cannot open annotation file $anno_file for writing: $!\n";

    my $assign_fh = new FileHandle(">$assign_file");
    $assign_fh or die "Cannot open assignment file $assign_file for writing: $!\n";
    
    for my $genome (@genomes)
    {
	print "Process $genome\n";
	walk_genome($anno_fh, $assign_fh, $genome);

	$anno_fh->autoflush(1);
	$anno_fh->autoflush(0);
	$assign_fh->autoflush(1);
	$assign_fh->autoflush(0);
    }

    close($assign_fh);
    close($anno_fh);
}


print "Global counts\n";
while (my ($k, $v) = each(%gcount))
{
    print "$k:\t$v\n";
}

sub walk_genome
{
    my($anno_fh, $assign_fh,  $genome) = @_;

    my($assigns, $assigns_mod) = load_assignments($genome);
    print "size ", int(keys %$assigns), " $assigns_mod\n";

    my $annos = load_latest_annos($genome);

    compare($anno_fh, $assign_fh, $genome, $assigns, $assigns_mod,  $annos);
}

sub compare
{
    my($anno_fh, $assign_fh, $genome, $assigns, $assigns_mod, $annos) = @_;

    my @pegs = sort { &FIG::by_fig_id($a, $b) } $fig->pegs_of($genome);

    my %counts;
    for my $peg (@pegs)
    {
	my $assign = $assigns->{$peg};
	my $anno = $annos->{$peg};

	my($fid, $anno_time, $who, $anno_text, $anno_fn);
	defined($anno) and ($fid, $anno_time, $who, $anno_text, $anno_fn) = @$anno;

	if (not defined($assign) and not defined($anno))
	{
	    #print "$peg: nothing\n";
	    $counts{none}++;
	}
	elsif (defined($assign) and not defined($anno))
	{
	    #print "$peg: no anno, $assign\n";
	    $counts{no_anno}++;
	    write_anno($anno_fh, $peg,  $assign, $assigns_mod, "Annotation for existing assignment\n$assign");
	}
	elsif (not defined($assign) and defined($anno))
	{
	    #print "$peg: no assign, $who $anno_fn\n";
	    write_assign($assign_fh, $peg, $anno_fn);
	    $counts{no_assign}++;
	}
	elsif ($assign eq $anno_fn)
	{
	    #print "$peg: OK\n";
	    $counts{ok}++;
	}
	else
	{
	    my $xx = $anno_fn;
	    $xx =~ s/\s+/ /g;
	    if ($xx eq $assign)
	    {
		write_anno($anno_fh, $peg, $assign, $assigns_mod, "Whitespace repair");
		#print "$peg: WS\n";
		$counts{ws}++;
	    }
	    else
	    {
		write_anno($anno_fh, $peg, $assign, $assigns_mod, "Annotation for existing assignment\n$assign\nthat did not match latest annotation\n$anno_fn");
		#print "$peg: clash: $anno_fn $who\n" . ' ' x length($peg);
		#print     "         $assign\n";
		$counts{clash}++;
	    }
	}
    }
    while (my ($k, $v) = each(%counts))
    {
	print "$k:\t$v\n";
	$gcount{$k} += $v;
    }
}

sub write_assign
{
    my($fh, $peg, $assignment) = @_;

    print $fh "$peg\t$assignment\n";
}

sub write_anno
{
    my($fh, $peg, $assignment, $time, $txt) = @_;

    print $fh join("\n", $peg, $time, 'annotation_repair', $txt), "\n";
    print $fh "///\n";
    print $fh join("\n", $peg, $time, 'annotation_repair', "Set master function to", $assignment), "\n";
    print $fh "///\n";
}

sub load_assignments
{
    my($genome) = @_;

    my $fh = new FileHandle("<$FIG_Config::organisms/$genome/assigned_functions");

    my $fns = {};

    if (!$fh)
    {
	warn "Cannot open assigned functions for $genome: $!";
	return {};
    }

    while (<$fh>)
    {
	chomp;
	my($peg, $assign) = split(/\t/);

	if (not $fig->is_deleted_fid($peg))
	{
	    $fns->{$peg}= $assign;
	}
    }

    my $stat = stat($fh);

    close($fh);
    return $fns, $stat->mtime;
}

sub load_latest_annos
{
    my($genome) = @_;

    #
    # %annos is a hash from fid to [$fid, $time, $who, $anno, $fn]
    #
    my %annos;

    my $fh = new FileHandle("<$FIG_Config::organisms/$genome/annotations");

    if (!$fh)
    {
	warn "Cannot open annotations for $genome: $!\n";
	return  \%annos;
    }

    local $/ = "//\n";

    my ($fid, $anno_time, $who, $anno_text, $anno_who, $fn);

    while (my $ann = <$fh>)
    {
	chomp $ann;

	if ((($fid, $anno_time, $who, $anno_text, $anno_who, $fn) =
	     ($ann =~ /^(fig\|\d+\.\d+\.peg\.\d+)\n(\d+)\n(\S+)\n(Set\s+(FIG|master)\s+function\s+to[ \t]*\n(.*))/s)) and
		    not $fig->is_deleted_fid($fid))
	{
	    my $cur = $annos{$fid};
	    
	    if (not ref($cur) or ($cur->[1] < $anno_time))
	    {
		$fn =~ s/\n//g;
		my $anno = [$fid, $anno_time, $who, $anno_text, $fn];
		$annos{$fid} = $anno;
	    }
	}
    }
    close($fh);
    return \%annos;
}

MCS Webmaster
ViewVC Help
Powered by ViewVC 1.0.3