Skip to content

Commit

Permalink
Added option --stranded_illumina for stranded mRNA and Total RNA kit.…
Browse files Browse the repository at this point in the history 
… Addresses #127.
  • Loading branch information
FelixKrueger committed Jan 15, 2022
1 parent 1a129b7 commit 072ecf9
Showing 1 changed file with 28 additions and 10 deletions.
38 changes: 28 additions & 10 deletions trim_galore
View file
Original file line number Diff line number Diff line change
Expand Up @@ -26,16 +26,16 @@ use Cwd;

## This script is taking in FastQ sequences and trims them using Cutadapt

## last modified on 10 09 2021
## last modified on 15 01 2022 to add Illumina directional mRNA kit adapter (ACTGTCTCTTATACACATCT)

my $DOWARN = 1; # print on screen warning and text by default
BEGIN { $SIG{'__WARN__'} = sub { warn $_[0] if $DOWARN } };

my $trimmer_version = '0.6.7';
my $trimmer_version = '0.6.8';
my $cutadapt_version;
my $python_version;

my ($compression_path,$cores,$cutoff,$adapter,$stringency,$rrbs,$length_cutoff,$keep,$fastqc,$non_directional,$phred_encoding,$fastqc_args,$trim,$gzip,$validate,$retain,$length_read_1,$length_read_2,$a2,$error_rate,$output_dir,$no_report_file,$dont_gzip,$clip_r1,$clip_r2,$three_prime_clip_r1,$three_prime_clip_r2,$nextera,$small_rna,$path_to_cutadapt,$illumina,$max_length,$maxn,$trim_n,$hardtrim5,$clock,$polyA,$hardtrim3,$nextseq,$basename,$consider_already_trimmed,$umi_from_r2) = process_commandline();
my ($compression_path,$cores,$cutoff,$adapter,$stringency,$rrbs,$length_cutoff,$keep,$fastqc,$non_directional,$phred_encoding,$fastqc_args,$trim,$gzip,$validate,$retain,$length_read_1,$length_read_2,$a2,$error_rate,$output_dir,$no_report_file,$dont_gzip,$clip_r1,$clip_r2,$three_prime_clip_r1,$three_prime_clip_r2,$nextera,$stranded_illumina,$small_rna,$path_to_cutadapt,$illumina,$max_length,$maxn,$trim_n,$hardtrim5,$clock,$polyA,$hardtrim3,$nextseq,$basename,$consider_already_trimmed,$umi_from_r2) = process_commandline();
my $report_message; # stores result of adapter auto-detection

my @filenames = @ARGV;
Expand Down Expand Up @@ -483,6 +483,13 @@ unless (defined $adapter){
$adapter = 'AGATCGGAAGAGC';
$adapter_name = 'Illumina TruSeq, Sanger iPCR; user defined';
}
elsif($stranded_illumina){
# more information available here:
# "Trimming T-overhang options for the Illumina Stranded mRNA and Illumina Stranded Total RNA workflows"
# https://support.illumina.com/bulletins/2020/06/trimming-t-overhang-options-for-the-illumina-rna-library-prep-wo.html
$adapter = 'ACTGTCTCTTATA';
$adapter_name = 'Illumina stranded mRNA and Total RNA; user defined (added for v0.6.8)';
}
else{ # default
### If other -a and/or -a2 were given
if ($polyA){ # specialised PolyA trimming
Expand Down Expand Up @@ -1254,6 +1261,8 @@ sub trim{

if ($nextseq){
$quality_cutoff = $nextseq;
warn "setting cutoff to: $nextseq\n";
sleep(1);
}
else{
$quality_cutoff = "-q $cutoff";
Expand Down Expand Up @@ -2355,6 +2364,7 @@ sub process_commandline{
my $three_prime_clip_r1;
my $three_prime_clip_r2;
my $nextera;
my $stranded_illumina; # added on 15 Jan 2022; https://support.illumina.com/bulletins/2020/06/trimming-t-overhang-options-for-the-illumina-rna-library-prep-wo.html
my $small_rna;
my $illumina;
my $path_to_cutadapt;
Expand Down Expand Up @@ -2404,6 +2414,7 @@ sub process_commandline{
'three_prime_clip_R2=i' => \$three_prime_clip_r2,
'illumina' => \$illumina,
'nextera' => \$nextera,
'stranded_illumina' => \$stranded_illumina,
'small_rna' => \$small_rna,
'path_to_cutadapt=s' => \$path_to_cutadapt,
'max_length=i' => \$max_length,
Expand Down Expand Up @@ -2442,7 +2453,7 @@ sub process_commandline{
[powered by Cutadapt]
version $trimmer_version
Last update: 11 05 2020
Last update: 15 01 2022
VERSION
exit;
Expand Down Expand Up @@ -2649,7 +2660,7 @@ VERSION
}


if ($nextera and $small_rna or $nextera and $illumina or $illumina and $small_rna ){
if ($nextera and $small_rna or $nextera and $illumina or $illumina and $small_rna or $nextera and $stranded_illumina or $stranded_illumina and $illumina or $stranded_illumina and $small_rna){
die "You can't use several different adapter types at the same time. Make your choice or consider using -a and -a2\n\n";
}

Expand Down Expand Up @@ -2702,13 +2713,13 @@ VERSION
}

if ($illumina){
die "You can't supply an adapter sequence AND use the Illumina universal adapter sequence. Make your choice.\n\n";
die "You can't supply an adapter sequence AND use the Illumina universal adapter sequence. Make your choice.\n\n";
}
if ($nextera){
die "You can't supply an adapter sequence AND use the Nextera transposase adapter sequence. Make your choice.\n\n";
die "You can't supply an adapter sequence AND use the Nextera transposase adapter sequence. Make your choice.\n\n";
}
if ($small_rna){
die "You can't supply an adapter sequence AND use the Illumina small RNA adapter sequence. Make your choice.\n\n";
die "You can't supply an adapter sequence AND use the Illumina small RNA adapter sequence. Make your choice.\n\n";
}
}

Expand Down Expand Up @@ -2989,7 +3000,7 @@ VERSION
$umi_from_r2 = 8;
}

return ($compression_path,$cores,$quality,$adapter,$stringency,$rrbs,$length_cutoff,$keep,$fastqc,$non_directional,$phred_encoding,$fastqc_args,$trim,$gzip,$validate,$retain,$length_read_1,$length_read_2,$adapter2,$error_rate,$output_dir,$no_report_file,$dont_gzip,$clip_r1,$clip_r2,$three_prime_clip_r1,$three_prime_clip_r2,$nextera,$small_rna,$path_to_cutadapt,$illumina,$max_length,$maxn,$trim_n,$hardtrim5,$clock,$polyA,$hardtrim3,$nextseq,$basename,$consider_already_trimmed,$umi_from_r2);
return ($compression_path,$cores,$quality,$adapter,$stringency,$rrbs,$length_cutoff,$keep,$fastqc,$non_directional,$phred_encoding,$fastqc_args,$trim,$gzip,$validate,$retain,$length_read_1,$length_read_2,$adapter2,$error_rate,$output_dir,$no_report_file,$dont_gzip,$clip_r1,$clip_r2,$three_prime_clip_r1,$three_prime_clip_r2,$nextera,$stranded_illumina,$small_rna,$path_to_cutadapt,$illumina,$max_length,$maxn,$trim_n,$hardtrim5,$clock,$polyA,$hardtrim3,$nextseq,$basename,$consider_already_trimmed,$umi_from_r2);
}

sub Ncounter{
Expand Down Expand Up @@ -3068,6 +3079,13 @@ trim_galore [options] <filename(s)>
--illumina Adapter sequence to be trimmed is the first 13bp of the Illumina universal adapter
'AGATCGGAAGAGC' instead of the default auto-detection of adapter sequence.
--stranded_illumina Adapter sequence to be trimmed is the first 13bp of the Illumina stranded mRNA or Total
RNA adapter 'ACTGTCTCTTATA' instead of the default auto-detection of adapter sequence.
Note that this sequence resembles the Nextera sequence with an additional A from A-tailing.
Please also see https://github.com/FelixKrueger/TrimGalore/issues/127 or
https://support.illumina.com/bulletins/2020/06/trimming-t-overhang-options-for-the-illumina-rna-library-prep-wo.html
for further information. This sequence is currently NOT included in the adapter auto-detection.
--nextera Adapter sequence to be trimmed is the first 12bp of the Nextera adapter
'CTGTCTCTTATA' instead of the default auto-detection of adapter sequence.
Expand Down Expand Up @@ -3363,7 +3381,7 @@ Paired-end specific options:
'.unpaired_2.fq' output file. These reads may be mapped in single-end mode.
Default: 35 bp.
Last modified on 10 September 2021.
Last modified on 15 January 2022.
HELP
exit;
Expand Down

0 comments on commit 072ecf9

Please sign in to comment.