Perl script that finds cases where different alternative splices of the same gene were put into different families, but those alternative splice forms overlap a lot at the DNA level

treefam_QC11.pl

#!/usr/local/bin/perl

# 
# Perl script treefam_QC11.pl
# Written by Avril Coghlan (a.coghlan@ucc.ie)
# 5-Feb-09.
#
# This perl script finds cases where different alternative splices of the
# same gene were put into different families, but those alternative splice
# forms overlap a lot at the DNA level.
#
# The command-line format is:
# % perl <treefam_QC11.pl> <release> 
#   where <release> is the release of the TreeFam database to use.
#
#
#------------------------------------------------------------------#

# CHECK IF THERE ARE THE CORRECT NUMBER OF COMMAND-LINE ARGUMENTS:

$num_args                  = $#ARGV + 1;
if ($num_args != 1)
{
   print "Usage of treefam_QC11.pl\n\n";
   print "perl treefam_QC11.pl <release>\n";
   print "where <release> is the release of the TreeFam database to use.\n";
   print "For example, >perl -w treefam_QC11.pl 7\n";
   exit;
}

#------------------------------------------------------------------#

# DECLARE MYSQL USERNAME AND HOST:
use DBI;

# FIND WHICH RELEASE OF THE TREEFAM DATABASE TO USE:

$release                   = $ARGV[0];

$VERBOSE                   = 1;

#------------------------------------------------------------------#

$database                  = "dbi:mysql:treefam_".$release.":db.treefam.org:3308";
$dbh                       = DBI->connect("$database", 'anonymous', '') || return;

# READ IN THE GENE ID FOR EVERY TRANSCRIPT ID:

$GID                       = &read_geneids($dbh);

# CHECK IF THERE ARE CASES WHERE DIFFERENT TRANSCRIPTS OF THE SAME
# GENE WERE ADDED TO DIFFERENT FAMILIES, WHERE THOSE TRANSCRIPTS
# OVERLAP A LOT AT THE DNA LEVEL:

&read_genes_in_families($dbh,$GID);

# NOW DISCONNECT FROM THE DATABASE:

$rc                        = $dbh->disconnect();
$rc                        = "";

#------------------------------------------------------------------#

print STDERR "FINISHED.\n";

#------------------------------------------------------------------#

# READ THE GENE ID FOR EVERY TRANSCRIPT ID:

sub read_geneids
{
   my $dbh                 = $_[0];
   my $table_w;
   my $st;
   my $sth;
   my $rv;
   my @array;
   my $ID;
   my $GID;
   my %GID                 = ();

   $table_w                = 'genes';
   $st                     = "SELECT ID, GID from $table_w";
   $sth                    = $dbh->prepare($st) or die "Cannot prepare $st: $dbh->errstr\n";
   $rv                     = $sth->execute or die "Cannot execute the query: $sth->errstr";
   if ($rv >= 1)
   {
      while ((@array) = $sth->fetchrow_array)
      {
         $ID               = $array[0];
         $GID              = $array[1];
         if (!($GID{$ID}))
         {
            $GID{$ID}      = $GID;
         } 
         else
         {
            if ($GID{$ID} ne $GID) { print STDERR "ERROR: id $ID GID $GID GID(ID) $GID{$ID}\n"; exit;}
         }
      }
   }

   print STDERR "Read in gene ids\n";

   return(\%GID);

}

#------------------------------------------------------------------#

# CHECK IF TWO TRANSCRIPTS OVERLAP MUCH AT THE DNA LEVEL:

sub check_if_overlap
{
   my $dbh                 = $_[0];
   my $id1                 = $_[1];
   my $id2                 = $_[2];
   my $table_w;
   my $st;
   my $sth;
   my $rv;
   my @array;
   my $i;
   my $id;
   my $START;
   my $STOP;
   my $start1              = "none";
   my $start2              = "none";
   my $stop1               = "none";
   my $stop2               = "none";
   my @start1;
   my @start2;
   my @stop1;
   my @stop2;
   my $overlap             = 0;
   my $start_id1;
   my $stop_id1;
   my $start_id2;
   my $stop_id2;
   my $j;
   my $length;

   $table_w                = 'map';
   for ($i = 1; $i <= 2; $i++)
   {
      if    ($i == 1) { $id = $id1;}
      elsif ($i == 2) { $id = $id2;}
      $st                  = "SELECT START,STOP from $table_w WHERE ID=\'$id\'";
      $sth                 = $dbh->prepare($st) or die "Cannot prepare $st: $dbh->errstr\n";
      $rv                  = $sth->execute or die "Cannot execute the query: $sth->errstr";
      if ($rv >= 1)
      {
         while ((@array) = $sth->fetchrow_array)
         {
            $START         = $array[0];
            $STOP          = $array[1];
            if    ($i == 1) { $start1 = $START; $stop1 = $STOP;}
            elsif ($i == 2) { $start2 = $START; $stop2 = $STOP;}
         }
      }
   }
   if ($start1 eq 'none' || $stop1 eq 'none') { print STDERR "ERROR: do not know coords for $id1\n"; exit;}
   if ($start2 eq 'none' || $stop2 eq 'none') { print STDERR "ERROR: do not know coords for $id2\n"; exit;} 
   @start1                 = split(/\,/,$start1);
   @start2                 = split(/\,/,$start2);
   @stop1                  = split(/\,/,$stop1);
   @stop2                  = split(/\,/,$stop2);
   for ($i = 0; $i <= $start1; $i++)
   {
      $start_id1           = $start1[$i];
      $stop_id1            = $stop1[$i];
      # SEE IF THIS EXON OVERLAPS WITH ANY EXON IN $id2: 
      for ($j = 0; $j <= $start2; $j++)
      {
         $start_id2        = $start2[$j];
         $stop_id2         = $stop2[$j];
         if (($start_id1 >= $start_id2 && $start_id1 <= $stop_id2) ||
             ($stop_id1  >= $start_id2 && $stop_id1  <= $stop_id2) ||
             ($start_id2 >= $start_id1 && $start_id2 <= $stop_id1) ||
             ($stop_id2  >= $start_id1 && $stop_id2  <= $stop_id1))
         {
            if    ($start_id1 >= $start_id2 && $stop_id1 <= $stop_id2) { $length = $stop_id1 - $start_id1 + 1;}
            elsif ($start_id2 >= $start_id1 && $stop_id2 <= $stop_id1) { $length = $stop_id2 - $start_id2 + 1;}
            elsif ($start_id1 >= $start_id2 && $stop_id2 <= $stop_id1) { $length = $stop_id2 - $start_id1 + 1;}
            elsif ($start_id2 >= $start_id1 && $stop_id1 <= $stop_id2) { $length = $stop_id1 - $start_id2 + 1;}
            $overlap       = $overlap + $length;
         }
      }
   }     
    
   return($overlap);
}

#------------------------------------------------------------------#

# CHECK IF THERE ARE CASES WHERE DIFFERENT TRANSCRIPTS OF THE SAME GENE
# WERE ADDED TO DIFFERENT FAMILIES, BUT THE DIFFERENT TRANSCRIPTS OVERLAP
# A LOT AT THE DNA LEVEL:

sub read_genes_in_families
{
   my $dbh                 = $_[0];
   my $GID                 = $_[1];
   my $table_w;
   my $st;
   my $sth;
   my $rv;
   my @array;
   my $FLAG;
   my %FAMILY              = ();
   my $geneid;
   my %SEEN;
   my $families;
   my @families;
   my $i;
   my $j;
   my $family1;
   my $family2;
   my @temp;
   my $id1;
   my $id2;

   $table_w                = 'fam_genes';
   $st                     = "SELECT AC, ID, FLAG from $table_w"; 
   $sth                    = $dbh->prepare($st) or die "Cannot prepare $st: $dbh->errstr\n";
   $rv                     = $sth->execute or die "Cannot execute the query: $sth->errstr";
   if ($rv >= 1)
   {
      while ((@array) = $sth->fetchrow_array)    
      {
         $AC               = $array[0];  
         $ID               = $array[1];
         $FLAG             = $array[2];
         # GET THE GENE ID FOR $ID:
         if (!($GID->{$ID})) 
         { 
            # THERE ARE SOME IDs IN THE fam_genes TABLE THAT DO NOT APPEAR IN THE genes
            # TABLE BECAUSE THEY BELONGED TO AN OLD TREEFAM-A SEED FAMILY AND ARE NOT IN
            # THE LATEST GENE SET FOR THAT SPECIES LOADED INTO THE genes TABLE.
            print STDERR "WARNING: do not know gene id for $ID (AC $AC FLAG $FLAG)\n"; 
         }
         else
         {
            $geneid        = $GID->{$ID}; 
            # RECORD THE FAMILIES THAT THIS GENE APPEARS IN:
            if (!($SEEN{$AC."_".$ID}))
            {
               $SEEN{$AC."_".$ID} = 1;
               if (!($FAMILY{$geneid}))
               {
                  $FAMILY{$geneid} = $AC."_".$ID;
               }   
               else
               {
                  $FAMILY{$geneid} = $FAMILY{$geneid}.",".$AC."_".$ID;
               }
            }
         } 
      }
   }

   # CHECK WHETHER THERE ARE CASES WHERE DIFFERENT TRANSCRIPTS OF THE
   # SAME GENE APPEAR IN DIFFERENT FAMILIES, BUT THOSE TRANSCRIPTS
   # OVERLAP A LOT AT THE DNA LEVEL:

   foreach $geneid (keys %FAMILY)
   {
      $families            = $FAMILY{$geneid};
      @families            = split(/\,/,$families);
      # LOOK AT EACH PAIR OF TRANSCRIPTS, WHICH ARE IN DIFFERENT FAMILIES:
      for ($i = 0; $i <= $#families; $i++)
      {
         $family1          = $families[$i];
         @temp             = split(/_/,$family1);
         $family1          = $temp[0];
         $id1              = $temp[1];
         for ($j = 0; $j <= $#families; $j++)
         {
            $family2       = $families[$j];
            @temp          = split(/_/,$family2);
            $family2       = $temp[0];
            $id2           = $temp[1];
            if ($family1 ne $family2 && $id1 ne $id2)
            {
               # CHECK WHETHER $id1 AND $id2 OVERLAP A LOT AT THE DNA LEVEL:
               #xxxyyyzzz
               # xxxyyyzzz
               $overlap    = &check_if_overlap($dbh,$id1,$id2); #xxx
               print "xxx $id1 (family $family1) overlaps by $overlap with $id2 ($family2)\n";
            }
         }
      } 
   }
}

#------------------------------------------------------------------#