annotate alignTophat.pl @ 13:cc676ee7fe40 draft

Uploaded

#!/usr/bin/perl
use strict;
use warnings;
use Getopt::Long;

my $max_procs = 8;

my ($ma, $fastq, $fastqN, $mis,$findSens, $contaminants, $directional, $build_contaminant, $ref, $build_ref, $annotation, $dirBam, $dirBam_cont, $dirBed, $dirBedgraph, $outText, $outSimple, $random_w);

GetOptions (
"direction:s" => \$directional,
"random_w:s" => \$random_w,
"find:s" => \$findSens,
"fastq=s" => \$fastq,
"fastq_n=s" => \$fastqN, 
"ma=i" => \$ma,
"mis=i" => \$mis,
"contaminants=s" => \$contaminants,
"anno=s" => \$annotation,
"build_contaminant" => \$build_contaminant,
"ref=s" => \$ref,
"build_ref" => \$build_ref,
"dirbamCont=s" => \$dirBam_cont,
"dirbam=s" => \$dirBam,
"dirbedgraph=s" => \$dirBedgraph,
"text=s" => \$outText,
);

$directional = lc($directional);
$findSens = lc($findSens);
$random_w = lc($random_w);

if ($build_ref)
{
  `(bowtie2-build $ref $ref) 2> /dev/null `;
}

if ($build_contaminant)
{
  `(bowtie2-build $contaminants $contaminants) 2> /dev/null `;
}

#"outsimple=s" => \$outSimple
#"annotation=s" => \$annotation,

#$annotation = "/data/oldgalaxy/annotation/ara/TAIR10_all_sorted.bed";
#$annotation = "/data/home/pogorelcnikr/SG/humanGenes.bed";

# mapping against contaminants databases  in: contaminants, fastq, # of mismatches, output_dir out: tophat dir
tophat2($fastq, $contaminants, $mis, 1, 'dir_contaminant');
`bamToFastq  -i dir_contaminant/unmapped.bam -fq clean.fastq`;

# mapping against contaminants databases  in: $ref, fastq, # of mismatches, output_dir out: tophat dir
tophat2('clean.fastq', $ref, $mis, $max_procs, 'dir_reference');

my $repTot = 0;
if ($ma == 0)
{
  open (my $rT,  'dir_reference/align_summary.txt') || die "cannot open dir_reference/align_summary.txt $!";
  my @repT = <$rT>;
  $repTot = $1 if $repT[2] =~ /\s+Mapped\s+:\s+(\d+).*/;
  close $rT;
}
else
{
  $repTot = $ma;
}

# Creation of bedgraphs
my $dir_bedgraph = $fastqN.'_bedgraph/';
findUnique('dir_reference/accepted_hits.bam', $dir_bedgraph, $repTot);

# count in: unique_mapped.bam out: raw_count_simple antisens_count
if ($random_w eq 'true')
{
  reference('rand_sorted.bam', $repTot);
  `mv rand_sorted.bam dir_reference/random_sorted.bam`;
}
else
{
  reference('ac_sorted.bam', $repTot);
  `mv ac_sorted.bam dir_reference/unique_sorted.bam`;
}

zip('dir_reference', $dirBam);
zip('dir_contaminant', $dirBam_cont);
zip( $dir_bedgraph , $dirBedgraph);

#################################
# inputs:
# fastq $_[0]
# bowtie2_index $_[1]
# mismatches # $_[2]
# output directory $_[3]
# max_procs $_[4]
#################################
sub tophat2
{
  my ($fastq, $index, $mismtaches, $proc, $out_repertory) = @_;
  $out_repertory = $out_repertory.'/' unless $out_repertory =~ /\/$/;
  # print STDERR "/data/visiteur/tophat-2.0.12.Linux_x86_64/tophat2 -p $proc -N $mismtaches --b2-very-sensitive -o $out_repertory $index $fastq\n";
  `tophat2 -p $proc -N $mismtaches --b2-very-sensitive -o $out_repertory $index $fastq 2>&1`;
}
############################################################################################

#################################
# inputs:
# bam $_[0]
# output directory $_[1]
#################################
sub findUnique
{
  my ($bam, $out_bedgraph, $sc) = @_;
  
  mkdir $out_bedgraph;

  my $bedg_uni =  $out_bedgraph.'/'.$fastqN.'_unique.bedgraph' ;
  my $bedg_all_multi = $out_bedgraph.'/'.$fastqN.'_all_multi.bedgraph' ;
  my $bedg_random = $out_bedgraph.'/'.$fastqN.'_random_multi.bedgraph' ;

  my $bedg_uni_p =  $out_bedgraph.'/'.$fastqN.'_unique_plusStrand.bedgraph' ;
  my $bedg_all_multi_p = $out_bedgraph.'/'.$fastqN.'_all_multi_plusStrand.bedgraph' ;
  my $bedg_random_p = $out_bedgraph.'/'.$fastqN.'_random_multi_plusStrand.bedgraph' ;

  my $bedg_uni_m =  $out_bedgraph.'/'.$fastqN.'_unique_minusStrand.bedgraph' ;
  my $bedg_all_multi_m = $out_bedgraph.'/'.$fastqN.'_all_multi_minusStrand.bedgraph' ;
  my $bedg_random_m = $out_bedgraph.'/'.$fastqN.'_random_multi_minusStrand.bedgraph' ;
  
  my $sca = 1000000/$sc;

  #unique file
  `samtools view -H $bam > ac.sam 2> /dev/null && samtools view -@ $max_procs $bam 2> /dev/null |grep NH:i:1\$ >>  ac.sam && samtools view -@ $max_procs -Shb ac.sam > ac.bam 2> /dev/null`;
  `samtools sort -@ $max_procs ac.bam -o ac_sorted.bam 2> /dev/null`;
  `genomeCoverageBed -scale $sca -split -bg -ibam ac_sorted.bam > $bedg_uni 2> /dev/null`;
  if ($directional eq 'true')
  {
    `genomeCoverageBed -strand + -scale $sca -split -bg -ibam ac_sorted.bam > $bedg_uni_p 2> /dev/null`;
    `genomeCoverageBed -strand - -scale $sca -split -bg -ibam ac_sorted.bam > $bedg_uni_m 2> /dev/null`;
  }
  #all multi
  `samtools sort -@ $max_procs $bam -o accepted_hits_sorted.bam 2> /dev/null`;
  `genomeCoverageBed -scale $sca -split -bg -ibam accepted_hits_sorted.bam > $bedg_all_multi 2> /dev/null`;
  if ($directional eq 'true')
  {
    `genomeCoverageBed -strand + -scale $sca -split -bg -ibam accepted_hits_sorted.bam > $bedg_all_multi_p 2> /dev/null`;
    `genomeCoverageBed -strand - -scale $sca -split -bg -ibam accepted_hits_sorted.bam > $bedg_all_multi_m 2> /dev/null`;
  }
  #random multi
  `samtools view -@ $max_procs -hb -F256 $bam > rand.bam 2> /dev/null`;
  `samtools sort -@ $max_procs rand.bam -o rand_sorted.bam 2> /dev/null`;
  `genomeCoverageBed -scale $sca -split -bg -ibam rand_sorted.bam > $bedg_random 2> /dev/null`;  
  if ($directional eq 'true')
  {
    `genomeCoverageBed -strand + -scale $sca -split -bg -ibam rand_sorted.bam > $bedg_random_p 2> /dev/null`;
    `genomeCoverageBed -strand - -scale $sca -split -bg -ibam rand_sorted.bam > $bedg_random_m 2> /dev/null`;
  }
}
############################################################################################

#################################
# inputs:
# bam $_[0]
# output directory $_[1]
#################################
sub reference
{
  my $bam = shift;
  my $mapnum = shift;  

  my @tabRef = ();
  
  open (my $anno, $annotation) || die "cannot open annotation file!\n";
  while(<$anno>)
  {
    chomp $_;
    my @split = split /\t/, $_;
    my $size = $split[2] - $split[1];
    push @tabRef, [$_,0,0,$size];
  }
  close $anno;
  
  `bedtools bamtobed -i $bam > ac.bed`;
  
 # my $mapnum = 0;
 # open (my $bed, 'ac.bed') || die "cannot open ac.bed\n";
 # while(<$bed>){ $mapnum++; }
 # close $bed;
  
  if ($directional eq 'true')
  {
    #Same strand
    `bedtools intersect -s -wao -b ac.bed -a $annotation> intersectStranded.bed`;
    #Different strand
    `bedtools intersect -S -wao -b ac.bed -a $annotation> intersectOppositeStrand.bed`;
  
    open (my $strand, 'intersectStranded.bed') || die "cannot open intersectStranded.bed\n";
    my $i = 0;
    my $firstLine = <$strand>; chomp $firstLine;
    my @split = split /\t/, $firstLine;
    my $prevId = $split[3];
    if ($split[16] != 0 && $split[12] - $split[11] == $split[16])
    {
      $tabRef[$i]->[1] = 1;
    }
    while(<$strand>)
    {
      chomp $_;
      @split = split /\t/, $_;
      if ($split[3] ne $prevId)
      {
        $i++;
      }
      if ($split[16] != 0 && $split[12] - $split[11] == $split[16])
      {
        $tabRef[$i]->[1] += 1;
      }
      $prevId = $split[3];
    }
    close $strand;
  
    open ($strand, 'intersectOppositeStrand.bed') || die "cannot open intersectOppositeStrand.bed\n";
    $i = 0;
    $firstLine = <$strand>; chomp $firstLine;
    @split = split /\t/, $firstLine;
    $prevId = $split[3];
    if ($split[16] != 0 && $split[12] - $split[11] == $split[16])
    {
      $tabRef[$i]->[2] = 1;
    }
    while(<$strand>)
    {
      chomp $_;
      @split = split /\t/, $_;
      if ($split[3] ne $prevId)
      {
        $i++;
      }
      if ($split[16] != 0 && $split[12] - $split[11] == $split[16])
      {
        $tabRef[$i]->[2] += 1;
      }
      $prevId = $split[3];
    }
    close $strand;
  }

  else
  {
    `bedtools intersect -wao -b ac.bed -a $annotation> intersect.bed`;

    open (my $strand, 'intersect.bed') || die "cannot open intersect.bed\n";
    my $i = 0;
    my $firstLine = <$strand>; chomp $firstLine;
    my @split = split /\t/, $firstLine;
    my $prevId = $split[3];
    if ($split[16] != 0 && $split[12] - $split[11] == $split[16])
    {
      $tabRef[$i]->[1] = 1;
    }
    while(<$strand>)
    {
      chomp $_;
      @split = split /\t/, $_;
      if ($split[3] ne $prevId)
      {
        $i++;
      }
      if ($split[16] != 0 && $split[12] - $split[11] == $split[16])
      {
        $tabRef[$i]->[1] += 1;
      }
      $prevId = $split[3];
    }
    close $strand; 
  }
  
  my $totplus = 0; my $totminus = 0;
  foreach my $outPut (@tabRef)
  {
    $totplus += $outPut->[1];
    $totminus += $outPut->[2];
  }
    
  open (my $out, ">".$outText) || die "cannot open raw_count.txt\n";
  print $out "Chr\tstart\tend\tID\tscore\tstrand\ttype\tD1\tD2\tD3\treads\trpm\treads on opposite strand\trpm\n";
  my ($Sr, $Rr) = (0,0);
  foreach my $outPut (@tabRef)
  {
    if ($totplus > $totminus || $findSens eq 'false')
    {
      $Sr = rpm($outPut->[1],$mapnum);
      $Rr = rpm($outPut->[2],$mapnum);
      print $out $outPut->[0]."\t".$outPut->[1]."\t".$Sr."\t".$outPut->[2]."\t".$Rr."\n";
    }
    else
    {
      $Rr = rpm($outPut->[1],$mapnum);
      $Sr = rpm($outPut->[2],$mapnum);
      print $out $outPut->[0]."\t".$outPut->[2]."\t".$Sr."\t".$outPut->[1]."\t".$Rr."\n";
    }
  }
  close $out;
}
############################################################################################

#################################
# inputs:
# bam $_[0]
# output directory $_[1]
#################################
sub rpm
{
  my $raw_count = shift;
  my $mapped_number = shift;
  my $RPM = 0;
  $RPM = ($raw_count * 1000000) / ($mapped_number) if  $mapped_number != 0;
  return $RPM;
}
############################################################################################

#################################
# inputs:
# bam $_[0]
# output directory $_[1]
#################################
sub zip
{
  my ($in_dir, $out_tar) = @_;
  `tar -cvf tmp.tar $in_dir/`;
  `mv tmp.tar $out_tar`;
  rmdir $in_dir;
}
############################################################################################