convert-sam-for-rsem

#!/usr/bin/env perl

use Getopt::Long;
use Pod::Usage;
use File::Basename;

use FindBin;
use lib $FindBin::RealBin;
use rsem_perl_utils qw(runCommand getSAMTOOLS);

use Env qw(@PATH);
@PATH = ($FindBin::RealBin, "$FindBin::RealBin/" . getSAMTOOLS(), @PATH);

use strict;


my $p = 1;
my $mem = "1G";
my $help = 0;

GetOptions("p|num-threads=i" => \$p,
	   "memory-per-thread=s" => \$mem,
	   "h|help" => \$help) or pd2usage(-exitval => 2, -verbose => 2);

	   
pod2usage(-verbose => 2) if ($help == 1);
pod2usage(-msg => "Invalid number of arguments!", -exitval => 2, -verbose => 2) if (scalar(@ARGV) != 2);

my $command;

my $in_file = $ARGV[0];
my $tmp_file = "$ARGV[1].tmp.bam";
my $out_file = "$ARGV[1].bam";

# Phase I, sort entries so that all entries of a same read groups together

$command = "samtools sort -n -@ $p -m $mem -o $tmp_file $in_file";
&runCommand($command);

# Phase II, parse the temporary BAM file to make paired-end alignments' two mates adjacent to each other

$command = "rsem-scan-for-paired-end-reads $p $tmp_file $out_file";
&runCommand($command);

# delete temporary file
unlink $tmp_file or die "Could not delete $tmp_file.";

print "Conversion is completed. $out_file will be checked by 'rsem-sam-validator'.\n";

# Phase III, validate if the resulting bam file is correct

$command = "rsem-sam-validator $out_file";
&runCommand($command);

__END__

=head1 NAME

convert-sam-for-rsem

=head1 PURPOSE

Make a RSEM compatible BAM file.

=head1 SYNOPSIS

 convert-sam-for-rsem [options] <input.sam/input.bam/input.cram> output_file_name

=head1 ARGUMENTS

=over

=item B<input.sam/input.bam/input.cram>

The SAM/BAM/CRAM file generated by user's aligner. We require this file contains the header section.

=item B<output_file_name>

The output name for the converted file. 'convert-sam-for-rsem' will output a BAM with the name 'output_file_name.bam'.

=back

=head1 OPTIONS

=over

=item B<-p/--num-threads> <int>

Set the number of threads to be used for converting. (Default: 1)

=item B<--memory-per-thread> <string>

Set the maximum allowable memory per thread. <string> represents the memory and accepts suffices 'K/M/G'. (Default: 1G)

=item B<-h/--help>

Show help information.

=back

=head1 DESCRIPTION

This program converts the SAM/BAM/CRAM file generated by user's aligner into a BAM file which RSEM can process. However, users should make sure their aligners use 'reference_name.idx.fa' generated by 'rsem-prepare-reference' as their references and output header sections. After the conversion, this program will call 'rsem-sam-validator' to validate the resulting BAM file. 

Note: You do not need to run this script if `rsem-sam-validator' reports that your SAM/BAM/CRAM file is valid.

=head1 EXAMPLES

Suppose input is set to 'input.sam' and output file name is "output" 

 convert-sam-for-rsem input.sam output

We will get a file called 'output.bam' as output.

=cut