annotate bin/align.pm @ 22:b9cf69322742 draft

Uploaded

package align;

use strict;
use warnings;
use File::Basename;
use String::Random;

use FindBin;
use lib $FindBin::Bin;
use Rcall qw ( histogram );

use Exporter;
our @ISA     = qw( Exporter );
our @EXPORT  = qw( &BWA_call &to_build &get_unique &sam_sorted_bam &get_hash_alignment &sam_to_bam_bg &sam_count &sam_count_mis &rpms_rpkm &get_fastq_seq &extract_sam );

sub to_build
{
	my $toBuildHashP  = shift; my $log = shift;

	while ( my ( $k, $v ) = each %{ $toBuildHashP } )
	{
		build_index ( $k, $log ) if $v == 1;
	}
}

sub build_index
{
	my $to_index = shift;
	my $log = shift;
	my $index_log = $to_index.'_index.err';
	`bwa index $to_index 2> $index_log`;
	print $log "Creating index for $to_index\n";
}

sub get_unique
{
	my ( $sam, $s_uni, $prefix, $details, $report ) = @_;

	my $fout = $prefix.'all.fastq'; 
	my $funi = $prefix.'unique.fastq';
	my $frej = $prefix.'rejected.fastq';
	
  my $repartition = $prefix.'distribution.txt';
	my $png_rep = $prefix.'distribution.png';
	my ( %duplicates, %genome_hits) ;

	#alignement to the first reference
	my @return = sam_parse( $sam, $fout, $funi, $frej, $s_uni, \%duplicates, \%genome_hits, $report );
	my $ref_fai = $return[4];
	my $mappers =  $return[5];
	my $mappers_uni = $return[6];
	my $size_mappedHashR = $return[7];

	if ( $details == 1 )
	{
		#print number of duplicates and hits number
		my ($pourcentage, $total) =(0,0);

		$total += $_ foreach values %{$size_mappedHashR};
		open (my $rep, '>'.$repartition) || die "cannot create $repartition $!\n";
		print $rep "size\tnumber\tpercentage\n";
		foreach my $k (sort{$a cmp $b} keys (%{$size_mappedHashR}))
		{
			$pourcentage = 0;
			$pourcentage = $size_mappedHashR->{$k} / $total * 100 unless $total ==0;

			print $rep "$k\t$size_mappedHashR->{$k}\t";
			printf $rep "%.2f\n",$pourcentage;
		}

		histogram($size_mappedHashR, $png_rep, $total);


		my $dup = $prefix.'dup_mapnum.txt';
		my $dup_u = $prefix .'dup_unique.txt';
		my $dup_r = $prefix .'dup_nonmapp.txt';
		open(my $tab,">".$dup) || die "cannot open output txt file\n";
		open(my $tab_r,">".$dup_r) || die "cannot open output txt file\n";
		open(my $tab_u,">".$dup_u) || die "cannot open output txt file\n";
		print $tab "sequence\tcount\tmapnum\n";
		print $tab_u "sequence\tcount\n";
		print $tab_r "sequence\tcount\n";
		foreach my $k (sort {$duplicates{$b} <=> $duplicates{$a}}keys %duplicates)
		{
			$duplicates{$k} = 0 unless exists($duplicates{$k});
			$genome_hits{$k} = 0 unless exists($genome_hits{$k});
			if ($genome_hits{$k} != 0) { print $tab $k."\t".$duplicates{$k}."\t".$genome_hits{$k}."\n"; }
			else {print $tab_r $k."\t".$duplicates{$k}."\n";}
			if ($genome_hits{$k} == 1) { print $tab_u $k."\t".$duplicates{$k}."\n"; }
		}
	close $dup; close $dup_r; close $dup_u;
	}
	return ( $ref_fai, $mappers, $mappers_uni );
}

sub sam_parse
{
	my ( $sam, $fastq_accepted, $fastq_accepted_unique, $fastq_rejected, $sam_unique, $duplicate_hashR, $best_hit_number_hashR, $report ) = @_ ;
	my ($reads, $mappers, $mappersUnique, @garbage, %size_num, %size_num_spe, %number, %numberSens, %numberReverse, %unique_number, %numberNM, %numberM, %size);
	$mappers = $mappersUnique = $reads = 0;

	open my $fic, '<', $sam || die "cannot open $sam $!\n";
	open my $accepted, '>', $fastq_accepted || die "cannot create $fastq_accepted $! \n";
	open my $unique, '>', $fastq_accepted_unique || die "cannot create $fastq_accepted_unique $! \n";
	open my $rejected, '>', $fastq_rejected || die "cannot create $fastq_rejected $! \n";
	open my $sam_uni, '>', $sam_unique || die "cannot create $sam_unique $! \n";
	my $sequence = '';
	while(<$fic>)
	{
		chomp $_;
		if ($_ =~ /^\@[A-Za-z][A-Za-z](\t[A-Za-z][A-Za-z0-9]:[ -~]+)+$/ || $_ =~ /^\@CO\t.*/ )
		{
			if ($_ =~ /\@SQ\tSN:(.*)\tLN:(\d*)/)
			{
				$size{$1} = $2;
				$unique_number{$1} = 0;
				$number{$1} = 0;
				$numberNM{$1} = 0;
				$numberM{$1} = 0;
			}
			print $sam_uni $_."\n";
			next;
		}
		$reads++;
		my @line = split (/\t/,$_);
		$sequence = $line[9];
		if ($line[1] & 16)
		{
			$sequence =reverse($sequence);
			$sequence =~ tr/atgcuATGCU/tacgaTACGA/;
		}
		if ($line[1] == 16 || $line[1] == 0)
		{
			my $len = length($sequence);
			$size_num{$len} ++;
			$size_num_spe{$line[2]}{$len}++;
			$mappers ++;

			${$best_hit_number_hashR}{$sequence} = $1  if  ($line[13] =~ /X0:i:(\d*)/ ||  $line[14] =~/X0:i:(\d*)/ );
			${$duplicate_hashR}{$sequence}++;
			$number{$line[2]}++;
			$numberSens{$line[2]}++ if $line[1] == 0 ;
			$numberReverse{$line[2]}++ if $line[1] == 16 ;
			print $accepted "\@".$line[0]."\n".$sequence."\n+\n".$line[10]."\n";

			if ($line[11] eq "XT:A:U")
			{
				$unique_number{$line[2]}++;
				$mappersUnique++;
				print $unique "\@".$line[0]."\n".$sequence."\n+\n".$line[10]."\n";
				print $sam_uni $_."\n";
			}
			if ($_ =~ /.*XM:i:(\d+).*/)
			{
				if ($1 == 0){$numberNM{$line[2]}++;}else{$numberM{$line[2]}++;}
			}
		}
		else
		{
			${$best_hit_number_hashR}{$sequence} = 0;
			${$duplicate_hashR}{$sequence}++;
			print $rejected "\@".$line[0]."\n".$sequence."\n+\n".$line[10]."\n";
		}
	}
	close $fic; close $accepted; close $unique; close $rejected; close $sam_uni;

	print $report "Parsing $sam file\n";
  print $report "\treads: $reads\n";	
	print $report "\tmappers: $mappers\n";
	print $report "\tunique mappers: $mappersUnique\n";
	print $report "-----------------------------\n";
	return (\%number, \%unique_number, \%numberSens, \%numberReverse, \%size, $mappers, $mappersUnique, \%size_num, \%size_num_spe, \%numberNM, \%numberM );
}

sub get_hash_alignment
{
  my ($index, $mismatches, $accept, $reject, $outA, $outR, $fastq, $number_of_cpus, $name, $sam, $report, $fai_f) = @_ ;
  my ($reads, $mappers, $unmapped) = (0,0,0);
  my $accep_unique;
  BWA_call ( $index, $fastq, $sam, $mismatches, $number_of_cpus, $report );
  
  open my $fic, '<', $sam || die "cannot open $sam $!\n";
  open my $accepted, '>', $outA || die "cannot open $outA\n"  if $accept == 1;
  open my $rejected, '>', $outR || die "cannot open $outR\n"  if $reject == 1; 
  open my $fai, '>', $fai_f || die "cannot open $fai_f\n" if $fai_f;
  #if ($name eq "snRNAs") {
  #  open ( $accep_unique, ">".$1."-unique.fastq") if $outR =~ /(.*)\.fastq/;
  #} 
  my $sequence = '';
  while(<$fic>)
  {
    chomp $_;
    if( $_ =~ /^\@[A-Za-z][A-Za-z](\t[A-Za-z][A-Za-z0-9]:[ -~]+)+$/ || $_ =~ /^\@CO\t.*/ )
    {
    	if ($fai_f && $_ =~ /\@SQ\tSN:(.*)\tLN:(\d*)/)
      {
        print $fai $1."\t".$2."\n";
      }
      next;
    }
    $reads++;
    my @line = split (/\t/,$_);
    $sequence = $line[9];
    if ($line[1] & 16)
    {
      $sequence =reverse($sequence);
      $sequence =~ tr/atgcuATGCU/tacgaTACGA/;
    }
    if ($line[1] & 16 || $line[1] == 0)
    {
      $mappers ++;
      if ($accept == 1 )
      {
        print $accepted "\@".$line[0]."\n".$sequence."\n+\n".$line[10]."\n";
   #     print $accep_unique "\@".$line[0]."\n".$sequence."\n+\n".$line[10]."\n" if ($name eq "snRNAs" && $line[11] eq "XT:A:U"); 
      }
    }
    else
    {
      print $rejected "\@".$line[0]."\n".$sequence."\n+\n".$line[10]."\n" if $reject == 1;
      $unmapped++;
    }
  }
 # close $accep_unique if ($name eq "bonafide_reads");
  close $fic;
  close $accepted  if $accept == 1;
  close $rejected if $reject ==1;
  close $fai if $fai_f;
  print $report "\treads: $reads\n";
  print $report "\tmappers: $mappers\n";
  print $report "\tunmapped: $unmapped\n";
  print $report "-----------------------------\n";
  return ($mappers, $unmapped);
}

sub sam_count
{
  my $sam = shift;
  my ( %number, %size );
  
  open  my $fic, '<', $sam || die "cannot open $sam file $!\n"; 
  while(<$fic>)
  {
    chomp $_;
    if ($_ =~ /^\@[A-Za-z][A-Za-z](\t[A-Za-z][A-Za-z0-9]:[ -~]+)+$/ || $_ =~ /^\@CO\t.*/ )
    {
      if ($_ =~ /\@SQ\tSN:(.*)\tLN:(\d*)/)
      {
        $size{$1} = $2;
        $number{$1} = 0;
      }      
    }
    else
    {
  		my @line = split (/\t/,$_);
 	  	if ( $line[1]  & 16 || $line[1] == 0 )
 	  	{
 	  		$number{$line[2]}++;
 	  	}
 	 	}
	}
  close $fic;
  return ( \%number, \%size );
}

sub sam_count_mis 
{

	my $sam = shift;
	my ( %number, %numberNM, %numberM, %size);
	
	open  my $fic, '<', $sam || die "cannot open $sam file $!\n";
  while(<$fic>)
  {
		chomp $_;
    if ($_ =~ /^\@[A-Za-z][A-Za-z](\t[A-Za-z][A-Za-z0-9]:[ -~]+)+$/ || $_ =~ /^\@CO\t.*/ )
    {
      if ($_ =~ /\@SQ\tSN:(.*)\tLN:(\d*)/)
      {
				$size{$1} = $2;
				$number{$1} = 0;
				$numberNM{$1} = 0;
				$numberM{$1} = 0;
			}
		}
		else
		{
			my @line = split (/\t/,$_);
			if ( $line[1]  & 16 || $line[1] == 0 )
			{
				$number{ $line[2] }++;
    	 	if ($_ =~ /.*XM:i:(\d+).*/)
    	 	{
        	if ( $1 == 0 ){ $numberNM{$line[2]}++; } else { $numberM{$line[2]}++; } 
      	}
			}
		}
	}
	return (\%number, \%size, \%numberNM, \%numberM );
}

sub sam_to_bam_bg
{
	my ( $sam, $scale, $number_of_cpus ) = @_;
	my ( $bam_sorted, $bedgraphM, $bedgraphP, $view_err, $sort_err ) = ( '', '', '', '', '' );
	if ( $sam =~ /(.*?).sam$/ )
	{
		$bam_sorted = $1.'_sorted.bam';
		$bedgraphP= $1.'_plus.bedgraph';
		$bedgraphM = $1.'_minus.bedgraph';
		$view_err = $1.'_view.err';
		$sort_err = $1.'_sort.err';
	}	
	`samtools view -Shb   --threads $number_of_cpus $sam 2> $view_err | samtools sort  -O BAM --threads $number_of_cpus  /dev/stdin 2> $sort_err  > $bam_sorted`;
	`bedtools genomecov -scale $scale -strand + -bga -ibam $bam_sorted > $bedgraphP`;	
	`bedtools genomecov -scale $scale -strand - -bga -ibam $bam_sorted > $bedgraphM`;
}

sub sam_sorted_bam
{
	my ( $sam, $number_of_cpus ) = @_;
	my ( $bam_sorted, $view_err, $sort_err ) = ( '', '', '' );
	if ( $sam =~ /(.*?).sam$/ )
	{
		$bam_sorted = $1.'_sorted.bam';
		$view_err = $1.'_view.err';
		$sort_err = $1.'_sort.err';

	}
	`samtools view -Shb   --threads $number_of_cpus $sam 2> $view_err | samtools sort  -O BAM --threads $number_of_cpus  /dev/stdin  2> $sort_err  > $bam_sorted`;
}

sub BWA_call
{
	my ( $index, $fastq, $sam, $mismatches, $number_of_cpus, $report ) = @_;
	my ( $aln_err, $samse_err, $seq_num ) = ( $sam.'_aln.err', $sam.'_samse.err', 0 );
	print $report "-----------------------------\n";
	print $report "bwa aln -t $number_of_cpus -n $mismatches $index $fastq 2> $aln_err | bwa samse $index /dev/stdin $fastq 2> $samse_err > $sam\n";
	`bwa aln -t $number_of_cpus -n $mismatches $index $fastq 2> $aln_err | bwa samse $index /dev/stdin $fastq 2> $samse_err > $sam `;
}

sub rpms_rpkm
{
  my ( $counthashR, $sizehashR, $mapped, $out_file, $piRNA_number, $miRNA_number, $bonafide_number ) =@_;
  open(my $out, ">".$out_file) || die "cannot open normalized file $! \n";
  print $out "ID\treads counts\tRPKM";
  print $out "\tper million of piRNAs" if ($piRNA_number != 0);
  print $out "\tper million of miRNAs" if ($miRNA_number != 0);
  print $out "\tper million of bonafide reads" if ($bonafide_number != 0);
  print $out "\n";
  foreach my $k  ( sort keys %{$counthashR} )
  {
    my ($rpkm, $pirna, $mirna, $bonafide) = (0,0,0,0);
    
    $rpkm = ( $counthashR->{$k} * 1000000000) / ( $sizehashR->{$k} * $mapped) if ( $sizehashR->{$k} * $mapped) != 0 ;
    print $out $k."\t".$counthashR->{$k}."\t"; printf $out "%.2f",$rpkm;
    
    if ($piRNA_number != 0 )
    {
      $pirna = ( $counthashR->{$k}  * 1000000) / $piRNA_number;
      printf $out "\t%.2f",$pirna;
    }
    if ($miRNA_number != 0 )
    {
      $mirna = ( $counthashR->{$k}  * 1000000) / $miRNA_number;
      printf $out "\t%.2f",$mirna;
    }
    if ($bonafide_number != 0 )
    {
      $bonafide = ( $counthashR->{$k}  * 1000000) / $bonafide_number;
      printf $out "\t%.2f",$bonafide;
    }
    print $out "\n";
 	}
  close $out;
}

sub extract_sam
{
  my ( $hashRef, $sam_in, $sam_out, $sam_uni_out, $fastq_out, $fastq_uni_out ) = @_;
  
	open my $s_in, '<', $sam_in || die "cannot open $sam_in file $!\n";

  open my $f_out, '>', $fastq_out || die "cannot create $fastq_out $!\n"; 
  open my $f_uni_out, '>', $fastq_uni_out || die "cannot create $fastq_uni_out $!\n"; 
  
  open my $s_out, '>', $sam_out || die "cannot create $sam_out file $!\n" if defined ($hashRef);
  open my $s_uni_out, '>', $sam_uni_out || die "cannot create $sam_uni_out file $!\n";

  my $sequence = '';
  while(<$s_in>)
  {
    if ($_ =~ /^\@[A-Za-z][A-Za-z](\t[A-Za-z][A-Za-z0-9]:[ -~]+)+$/ || $_ =~ /^\@CO\t.*/ )
    {
      print $s_out $_ if defined ($hashRef);
      print $s_uni_out $_;
      next;
    }
    my @line = split (/\t/,$_);
    $sequence = $line[0];
    if ( (! defined ($hashRef) )|| (  exists $hashRef->{$sequence}  &&  $hashRef->{$sequence} == 1 ) )
    {
      my $arn  =  $line[9];
      if ($line[1] & 16)
      {
        $arn =reverse($arn);
        $arn =~ tr/atgcuATGCU/tacgaTACGA/;
      }

      if  ( ( $line[1] == 16 || $line[1] == 0 ) )
			{
      	print $f_out "\@".$line[0]."\n".$arn."\n+\n".$line[10]."\n" ;
      	print $s_out $_ if defined ($hashRef);
      	if ( $line[11] eq "XT:A:U" )
      	{
      		print $f_uni_out "\@".$line[0]."\n".$arn."\n+\n".$line[10]."\n" ;
      		print $s_uni_out $_ ;
      	}
      }
    } 
  }
  close $s_in; close $s_out if defined ($hashRef); 
  close $s_uni_out; close $f_out; close $f_uni_out;
}

sub get_fastq_seq{
  my $fastq = shift;
  my %hash; my $cmp = 0; 

  open my $fic, '<', $fastq || die "cannot open input file $! \n";
  while(<$fic>)
  {
    chomp $_;
    $cmp++;
    if ($cmp % 4 == 1)
    {
      die "file do not contain a @ at line $cmp\n" unless ($_ =~ /^\@/ );
      if ($_ =~ /^\@(.*)\s.*/) { $hash{$1} = 1;}
      elsif ($_ =~ /^\@(.*)/) { $hash{$1} = 1;}
    }
    elsif ($cmp % 4 == 3 )
    {
      die "file do not contain a + at line $cmp\n" unless $_ =~ /^\+/;
    }
  }
  close $fic;
  return \%hash;
}

1;