clean_gencode_annotation

#!/usr/bin/perl -w

use warnings;
use strict;
use POSIX;
use Getopt::Long;
use File::Path;
use File::Basename;
use Pod::Usage;

=head1 NAME

clean_gencode_annotation

=head1 SYNOPSIS

 clean_gencode_annotation [options]  
  
=head1 DESCRIPTION
 
 Removes unnecessary annotation.  
 
=cut

#option variables
my $help;
my $verbose;
my $debug;

#initialize options
Getopt::Long::Configure ('bundling');

if(!GetOptions ('h'=>\$help, 'v'=>\$verbose, 'd'=>\$debug) 
  || scalar(@ARGV)!=1)
{
    if ($help)
    {
        pod2usage(-verbose => 2);
    }
    else
    {
        pod2usage(1);
    }
}

my $gencodeAnnotationGTFFile = $ARGV[0];

open(GTF, "zcat $gencodeAnnotationGTFFile |") || die "Cannot open $gencodeAnnotationGTFFile";

#        //populate interval trees with reference sets
#        //gene transfer format
#        //1   chr1
#        //2   HAVANA - source
#        //3   exon   - feature
#        //4   13221  - start
#        //5   14409  - end
#        //6   .      - score
#        //7   +      - strand
#        //8   .      - frame
#        //9   gene_id "ENSG - attribute
#            //gene_id "ENSG00000149656.4";
#            //transcript_id "ENST00000425473.1";
#            //gene_type "processed_transcript";
#            //gene_status "KNOWN";
#            //gene_name "LINC00266-1";
#            //transcript_type "processed_transcript";
#            //transcript_status "KNOWN";
#            //transcript_name "LINC00266-1-002"; exon_number 3;
#            //exon_id "ENSE00002450675.1";
#            //level 2; havana_gene "OTTHUMG00000033036.2";
#            //havana_transcript "OTTHUMT00000080305.1";

while(<GTF>)
{
    s/\r?\n?$//;
    
    if (/^#/)
    {
        print "$_\n";    
    }
    else
    {
        my ($chrom, $src, $feature, $start, $end, $score, $strand, $frame, $attribute) = split("\t", $_, 10);
        
        my @ATTRIB = split(';', $attribute);
        my @newAttribs = ();
        
        for my $attrib (@ATTRIB)
        {
            $attrib =~ s/^\s+//;
            
            my @fields = split(' ', $attrib);
            
            if ($fields[0] eq "gene_type" ||
                $fields[0] eq "gene_name" ||
                $fields[0] eq "transcript_type" ||
                $fields[0] eq "level")
            {
                push(@newAttribs, $attrib);
            }
        }
        
        print "$chrom\t$src\t$feature\t$start\t$end\t$score\t$strand\t$frame\t" . join(';', @newAttribs) . "\n";
    
    }
}

close(GTF);