trim_galore: trim_galore.xml comparison

comparison trim_galore.xml @ 16:cd7e644cae1d draft default tip

"planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/trim_galore commit 276a0ec327f5369c16563696047f0d31577c353f"

author	bgruening
date	Fri, 08 Oct 2021 09:57:52 +0000
parents	084bbd8ba7b8
children

comparison

equal deleted inserted replaced

-:084bbd8ba7b8
+:cd7e644cae1d
-<tool id="trim_galore" name="Trim Galore!" version="0.6.3" profile="17.01">
+<tool id="trim_galore" name="Trim Galore!" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@" profile="20.01">
-<!-- Wrapper compatible with Trim Galore! version 0.6.3 -->
 <description>Quality and adapter trimmer of reads</description>
 <macros>
-<macro name="adapter_trimming">
+<import>macros.xml</import>
-<conditional name="trimming">
-<param name="trimming_select" type="select" label="Adapter sequence to be trimmed">
-<option value="">Automatic detection</option>
-<option value="--illumina">Illumina universal</option>
-<option value="--nextera">Nextera transposase</option>
-<option value="--small_rna">Illumina small RNA adapters</option>
-<option value="user">User defined adapter sequence</option>
-</param>
-<when value=""/>
-<when value="--illumina"/>
-<when value="--nextera"/>
-<when value="--small_rna"/>
-<when value="user">
-<param name="adapter" type="text" value="AGATCGGAAGAGC" label="Adapter sequence to be trimmed off">
-<validator type="regex" message="Adapter sequence must contain DNA characters only (A,C,T,G or N)">^[ACTGNactgn]*$</validator>
-</param>
-<yield/>
-</when>
-</conditional>
-</macro>
-<macro name="paired_adapter_trimming">
-<expand macro="adapter_trimming">
-<param name="adapter2" type="text" optional="True" value="" label="Adapter sequence to be trimmed off read 2">
-<validator type="regex" message="Adapter sequence must contain DNA characters only (A,C,T,G or N)">^[ACTGNactgn]*$</validator>
-</param>
-</expand>
-<param name="trim1" type="boolean" truevalue="--trim1" falsevalue="" checked="False" label="Trims 1 bp off every read from its 3' end." help="" />
-<param name="three_prime_clip_R1" type="integer" value="" optional="True" label="Remove N bp from the 3' end of read 1">
-<help>Instructs Trim Galore! to remove N bp from the 3' end of read 1 after adapter/quality trimming has been performed.
-This may remove some unwanted bias from the 3' end that is not directly related to adapter sequence or basecall quality.
-(--three_prime_clip_R1)</help>
-</param>
-<param name="three_prime_clip_R2" type="integer" value="" optional="True" label="Remove N bp from the 3' end of read 2">
-<help>Instructs Trim Galore! to remove N bp from the 3' end of read 2 after
-adapter/quality trimming has been performed. This may remove some unwanted bias from
-the 3' end that is not directly related to adapter sequence or basecall quality.</help>
-</param>
-</macro>
 </macros>
-<requirements>
+<expand macro="requirements" />
-<requirement type="package" version="0.6.3">trim-galore</requirement>
+<expand macro="xrefs"/>
-</requirements>
 <version_command>
 trim_galore --version
 </version_command>
 <command detect_errors="aggressive"><![CDATA[
 #set compressed = 'no'
 ln -s '${singlePaired.input_mate_pairs.reverse}' ${read2} &&
 #end if
 trim_galore
+## according the develpers 4 cores could be a sweet spot, anything above has diminishing returns
+--cores \${GALAXY_SLOTS:-4}
 ## we only support fastqsanger
 --phred33
 #if $params.settingsType == "custom":
 ${read2}
 #end if
 #if $compressed == 'no':
 --dont_gzip
+#end if
+## Trimming settings
+#if $trimming.settingsType == 'custom'
+#if $trimming.hardtrim5
+--hardtrim5 $trimming.hardtrim5
+#end if
+#if $trimming.hardtrim3
+--hardtrim3 $trimming.hardtrim3
+#end if
+$trimming.clock
+$trimming.polyA
 #end if
 ##  Trim Galore is finished, rename the output if compressed
 &&
 if [ -f input_1_trimmed.fq.gz ] ; then mv input_1_trimmed.fq.gz input_1_trimmed.fq ; fi
 if [ -f input_2_val_2.fq.gz ] ; then mv input_2_val_2.fq.gz input_2_val_2.fq ; fi
 &&
 if [ -f input_1_unpaired_1.fq.gz ] ; then mv input_1_unpaired_1.fq.gz input_1_unpaired_1.fq ; fi
 &&
 if [ -f input_2_unpaired_2.fq.gz ] ; then mv input_2_unpaired_2.fq.gz input_2_unpaired_2.fq ; fi
+&&
+if [ -f input_1.clock_UMI.R1.fq.gz ] ; then mv input_1.clock_UMI.R1.fq.gz input_1.clock_UMI.R1.fq ; fi
+&&
+if [ -f input_2.clock_UMI.R2.fq.gz ] ; then mv input_2.clock_UMI.R2.fq.gz input_2.clock_UMI.R2.fq ; fi
+## Rename hardtrimmed files
+#if $trimming.settingsType == 'custom'
+&&
+if [ -f input_1.${trimming.hardtrim5}bp_5prime.fq.gz ] ; then mv input_1.${trimming.hardtrim5}bp_5prime.fq.gz input_1_hardtrim.fq ; fi
+&&
+if [ -f input_2.${trimming.hardtrim5}bp_5prime.fq.gz ] ; then mv input_2.${trimming.hardtrim5}bp_5prime.fq.gz input_2_hardtrim.fq ; fi
+&&
+if [ -f input_1.${trimming.hardtrim3}bp_3prime.fq.gz ] ; then mv input_1.${trimming.hardtrim3}bp_3prime.fq.gz input_1_hardtrim.fq ; fi
+&&
+if [ -f input_2.${trimming.hardtrim3}bp_3prime.fq.gz ] ; then mv input_2.${trimming.hardtrim3}bp_3prime.fq.gz input_2_hardtrim.fq ; fi
+&&
+if [ -f input_1.${trimming.hardtrim5}bp_5prime.fq ] ; then mv input_1.${trimming.hardtrim5}bp_5prime.fq input_1_hardtrim.fq ; fi
+&&
+if [ -f input_2.${trimming.hardtrim5}bp_5prime.fq ] ; then mv input_2.${trimming.hardtrim5}bp_5prime.fq input_2_hardtrim.fq ; fi
+&&
+if [ -f input_1.${trimming.hardtrim3}bp_3prime.fq ] ; then mv input_1.${trimming.hardtrim3}bp_3prime.fq input_1_hardtrim.fq ; fi
+&&
+if [ -f input_2.${trimming.hardtrim3}bp_3prime.fq ] ; then mv input_2.${trimming.hardtrim3}bp_3prime.fq input_2_hardtrim.fq ; fi
+#end if
 ##  Trim Galore! run is finished. Move the report files to the proper place
 #if $params.settingsType == "custom" and $params.report:
 &&
 cat ./*_trimming_report.txt > '$report_file'
 #end if
+&& ls -lah
 ]]></command>
 <inputs>
 <!-- Input Parameters -->
 <conditional name="singlePaired">
 <param name="sPaired" type="select" label="Is this library paired- or single-end?">
 <expand macro="paired_adapter_trimming" />
 </when>
 </conditional>
 <conditional name="params">
-<param name="settingsType" type="select" label="Trim Galore! advanced settings" help="You can use the default settings or set custom values for any of Trim Galore!'s parameters.">
+<param name="settingsType" type="select" label="Advanced settings" help="You can use the default settings or set custom values for any of Trim Galore!'s parameters.">
 <option value="default">Use defaults</option>
 <option value="custom">Full parameter list</option>
 </param>
 <when value="default" />
 <!-- Full/advanced params. -->
 </when>  <!-- full -->
 </conditional>  <!-- params -->
 <conditional name="rrbs">
-<param name="settingsType" type="select" label="RRBS specific settings">
+<param name="settingsType" type="select" label="RRBS specific settings" help="You can use the default settings or set custom values for any of Trim Galore!'s parameters.">
 <option value="default">Use defaults (no RRBS)</option>
 <option value="custom">Full parameter list</option>
 </param>
 <when value="default" />
 <!-- Full/advanced params. -->
 label="Specifies that the input file was an MspI digested RRBS sample" />
 <param name="non_directional" type="boolean" truevalue="--non_directional" falsevalue="" checked="False"
 label="Screen quality-trimmed sequences for 'CAA' or 'CGA' at the start of the read and, if found, removes the first two basepairs" />
 </when>  <!-- full -->
 </conditional>  <!-- params -->
+<!--Trimming options-->
+<conditional name="trimming">
+<param name="settingsType" type="select" label="Trimming settings" help="You can use the default settings or set custom values for any of Trim Galore!'s parameters.">
+<option value="default">Use defaults</option>
+<option value="custom">Full parameter list</option>
+</param>
+<when value="default" />
+<when value="custom">
+<param argument="--hardtrim5" type="integer" min="1" optional="true" label="Hard-trimm 5' ends" help="Instead of performing adapter-/quality trimming, this option will simply hard-trim sequences to N bp at the 5'-end. Once hard-trimming of files is complete, it will exit" />
+<param argument="--hardtrim3" type="integer" min="1" optional="true" label="Hard-trimm 3' ends" help="Instead of performing adapter-/quality trimming, this option will simply hard-trim sequences to N bp at the 3'-end. Once hard-trimming of files is complete, it will exit" />
+<param argument="--clock" type="boolean" truevalue="--clock" falsevalue="" label="Mouse epigenetic clock mode" help="In this mode, reads are trimmed in a specific way that is currently used for the Mouse Epigenetic Clock"/>
+<param argument="--polyA" type="boolean" truevalue="--polyA" falsevalue="" label="Remove polyA tails" help="This is a new, still experimental, trimming mode to identify and remove poly-A tails from sequences" />
+</when>
+</conditional>
 </inputs>
 <outputs>
 <data format_source="input_singles" name="trimmed_reads_single" from_work_dir="input_1_trimmed.fq" label="${tool.name} on ${on_string}: trimmed reads">
 <filter>singlePaired['sPaired'] == "single"</filter>
-</data>
+<filter>trimming['hardtrim3'] == '' and trimming['hardtrim5'] == ''</filter>
+</data>
+<data format_source="input_singles" name="hardtrim_reads_single" from_work_dir="input_1_hardtrim.fq" label="${tool.name} on ${on_string}: hard-trimmed reads">
+<filter>singlePaired['sPaired'] == "single"</filter>
+<filter>trimming['settingsType'] == "custom"</filter>
+<filter>trimming['hardtrim3'] != '' or trimming['hardtrim5'] != ''</filter>
+</data>
+<!--Trimmed reads paired collection-->
 <collection name="trimmed_reads_paired_collection" type="paired" label="${tool.name} on ${on_string}: paired reads">
 <data name="forward" format_source="input_mate_pairs['forward']" from_work_dir="input_1_val_1.fq" />
 <data name="reverse" format_source="input_mate_pairs['forward']" from_work_dir="input_2_val_2.fq" />
 <filter>singlePaired['sPaired'] == "paired_collection"</filter>
+<filter>trimming['hardtrim3'] == '' and trimming['hardtrim5'] == ''</filter>
+<filter>trimming['clock'] == False</filter>
 </collection>
+<!--Unpaired reads collection-->
 <collection name="trimmed_reads_unpaired_collection" type="paired" label="${tool.name} on ${on_string}: unpaired reads">
 <data name="forward" format_source="input_mate_pairs['forward']" from_work_dir="input_1_unpaired_1.fq" />
 <data name="reverse" format_source="input_mate_pairs['forward']" from_work_dir="input_2_unpaired_2.fq" />
 <filter>params['settingsType'] == "custom"</filter>
 <filter>params['retain_unpaired']['retain_unpaired_select'] == "retain_unpaired_output"</filter>
 <filter>singlePaired['sPaired'] == "paired_collection"</filter>
+<filter>trimming['hardtrim3'] == '' and trimming['hardtrim5'] == ''</filter>
 </collection>
+<!--Hard-trimmed reads paired collection-->
+<collection name="hardtrimmed_reads_paired_collection" type="paired" label="${tool.name} on ${on_string}: hard-trimmed paired reads">
+<data name="forward" format_source="input_mate_pairs['forward']" from_work_dir="input_1_hardtrim.fq" />
+<data name="reverse" format_source="input_mate_pairs['forward']" from_work_dir="input_2_hardtrim.fq" />
+<filter>singlePaired['sPaired'] == "paired_collection"</filter>
+<filter>trimming['settingsType'] == "custom"</filter>
+<filter>trimming['hardtrim3'] or trimming['hardtrim5']</filter>
+</collection>
+<!--Mouse epigenetic reads paired collection-->
+<collection name="mouse_reads_paired_collection" type="paired" label="${tool.name} on ${on_string}: MEC paired reads">
+<data name="forward" format_source="input_mate_pairs['forward']" from_work_dir="input_1.clock_UMI.R1.fq" />
+<data name="reverse" format_source="input_mate_pairs['forward']" from_work_dir="input_2.clock_UMI.R2.fq" />
+<filter>singlePaired['sPaired'] == "paired_collection"</filter>
+<filter>trimming['settingsType'] == "custom"</filter>
+<filter>trimming['clock']</filter>
+</collection>
 <data format_source="input_mate1" name="trimmed_reads_pair1" from_work_dir="input_1_val_1.fq"
 label="${tool.name} on ${on_string}: trimmed reads pair 1">
 <filter>singlePaired['sPaired'] == "paired"</filter>
+<filter>trimming['hardtrim3'] == '' and trimming['hardtrim5'] == ''</filter>
+<filter>trimming['clock'] == False</filter>
 </data>
 <data format_source="input_mate2" name="trimmed_reads_pair2" from_work_dir="input_2_val_2.fq"
 label="${tool.name} on ${on_string}: trimmed reads pair 2">
 <filter>singlePaired['sPaired'] == "paired"</filter>
+<filter>trimming['hardtrim3'] == '' and trimming['hardtrim5'] == ''</filter>
+<filter>trimming['clock'] == False</filter>
 </data>
 <data format_source="input_mate1" name="unpaired_reads_1" from_work_dir="input_1_unpaired_1.fq"
 label="${tool.name} on ${on_string}: unpaired reads (1)">
 <filter>params['settingsType'] == "custom"</filter>
 label="${tool.name} on ${on_string}: unpaired reads (2)">
 <filter>params['settingsType'] == "custom"</filter>
 <filter>params['retain_unpaired']['retain_unpaired_select'] == "retain_unpaired_output"</filter>
 <filter>singlePaired['sPaired'] == "paired"</filter>
 </data>
+<!--Hard-trimmed paired reads-->
+<data format_source="input_mate1" name="hardtrimmed_reads_pair1" from_work_dir="input_1_hardtrim.fq"
+label="${tool.name} on ${on_string}: hard-trimmed reads pair 1">
+<filter>singlePaired['sPaired'] == "paired"</filter>
+<filter>trimming['settingsType'] == 'custom'</filter>
+<filter>trimming['hardtrim3'] or trimming['hardtrim5']</filter>
+</data>
+<data format_source="input_mate2" name="hardtrimmed_reads_pair2" from_work_dir="input_2_hardtrim.fq"
+label="${tool.name} on ${on_string}: hard-trimmed reads pair 2">
+<filter>singlePaired['sPaired'] == "paired"</filter>
+<filter>trimming['settingsType'] == 'custom'</filter>
+<filter>trimming['hardtrim3'] or trimming['hardtrim5']</filter>
+</data>
+<!--Mouse epigenetic mode paired reads-->
+<data format_source="input_mate1" name="mec_reads_pair1" from_work_dir="input_1.clock_UMI.R1.fq"
+label="${tool.name} on ${on_string}: MEC reads pair 1">
+<filter>singlePaired['sPaired'] == "paired"</filter>
+<filter>trimming['settingsType'] == 'custom'</filter>
+<filter>trimming['clock']</filter>
+</data>
+<data format_source="input_mate2" name="mec_reads_pair2" from_work_dir="input_2.clock_UMI.R2.fq"
+label="${tool.name} on ${on_string}: MEC reads pair 2">
+<filter>singlePaired['sPaired'] == "paired"</filter>
+<filter>trimming['settingsType'] == 'custom'</filter>
+<filter>trimming['clock']</filter>
+</data>
 <data format="txt" name="report_file" label="${tool.name} on ${on_string}: report file">
 <filter>params['settingsType'] == "custom"</filter>
 <filter>params['report'] == True</filter>
 </data>
 </outputs>
 <tests>
-<test>
+<test expect_num_outputs="2">
 <param name="input_singles" value="sanger_full_range_original_sanger.fastqsanger" ftype="fastqsanger" />
 <param name="sPaired" value="single" />
 <param name="settingsType" value="custom" />
 <param name="report" value="true" />
 <output name="trimmed_reads_single" file="sanger_full_range_results1.fastqsanger" ftype="fastqsanger"/>
-<output name="report_file" file="sanger_full_range_report_results1.txt" ftype="txt" lines_diff="8" />
+<output name="report_file" file="sanger_full_range_report_results1.txt" ftype="txt" lines_diff="12" />
 </test>
-<test>
+<test expect_num_outputs="2">
 <param name="input_singles" value="sanger_full_range_original_sanger.fastq.gz" ftype="fastqsanger.gz" />
 <param name="sPaired" value="single" />
 <param name="settingsType" value="custom" />
 <param name="report" value="true" />
 <output name="trimmed_reads_single" file="sanger_full_range_results1.fastq.gz" ftype="fastqsanger.gz" decompress="true" />
-<output name="report_file" file="sanger_full_range_report_results1gz.txt" ftype="txt" lines_diff="9" />
+<output name="report_file" file="sanger_full_range_report_results1gz.txt" ftype="txt" lines_diff="12" />
 </test>
-<test>
+<test expect_num_outputs="1">
 <param name="input_singles" value="sanger_full_range_original_sanger.fastqsanger" ftype="fastqsanger" />
 <param name="sPaired" value="single" />
 <param name="trimming_select" value="--illumina" />
 <output name="trimmed_reads_single" file="sanger_full_range_results2.fastqsanger" ftype="fastqsanger"/>
 </test>
-<test>
+<test expect_num_outputs="1">
 <param name="input_singles" value="sanger_full_range_original_sanger.fastq.gz" ftype="fastqsanger.gz" />
 <param name="sPaired" value="single" />
 <param name="trimming_select" value="--illumina" />
 <output name="trimmed_reads_single" file="sanger_full_range_results2.fastq.gz" ftype="fastqsanger.gz" decompress="true" />
 </test>
-<test>
+<test expect_num_outputs="1">
 <param name="input_singles" value="sanger_full_range_original_sanger.fastqsanger" ftype="fastqsanger" />
 <param name="sPaired" value="single" />
 <param name="adapter" value="AAAGAGC" />
 <output name="trimmed_reads_single" file="sanger_full_range_results3.fastqsanger" ftype="fastqsanger"/>
 </test>
-<test>
+<test expect_num_outputs="1">
 <param name="input_singles" value="sanger_full_range_original_sanger.fastq.gz" ftype="fastqsanger.gz" />
 <param name="sPaired" value="single" />
 <param name="adapter" value="AAAGAGC" />
 <output name="trimmed_reads_single" file="sanger_full_range_results3.fastq.gz" ftype="fastqsanger.gz" decompress="true" />
 </test>
-<test>
+<test expect_num_outputs="3">
 <param name="input_mate1" value="bwa-mem-fastq1.fq" ftype="fastqsanger" />
 <param name="input_mate2" value="bwa-mem-fastq2.fq" ftype="fastqsanger" />
 <param name="sPaired" value="paired" />
 <param name="settingsType" value="custom" />
 <param name="report" value="true" />
 <output name="trimmed_reads_pair1" file="paired_example_pair1_results2.fastqsanger" ftype="fastqsanger"/>
 <output name="trimmed_reads_pair2" file="paired_example_pair2_results2.fastqsanger" ftype="fastqsanger"/>
 <output name="report_file" file="paired_example_results2.txt" ftype="txt" lines_diff="24" />
 </test>
-<test>
+<test expect_num_outputs="3">
 <param name="input_mate1" value="bwa-mem-fastq1.fq.gz" ftype="fastqsanger.gz" />
 <param name="input_mate2" value="bwa-mem-fastq2.fq.gz" ftype="fastqsanger.gz" />
 <param name="sPaired" value="paired" />
 <param name="settingsType" value="custom" />
 <param name="report" value="true" />
 <output name="trimmed_reads_pair1" file="paired_example_pair1_results2.fastq.gz" ftype="fastqsanger.gz" decompress="true" />
 <output name="trimmed_reads_pair2" file="paired_example_pair2_results2.fastq.gz" ftype="fastqsanger.gz" decompress="true" />
 <output name="report_file" file="paired_example_results2gz.txt" ftype="txt" lines_diff="24" />
 </test>
-<test>
+<test expect_num_outputs="7">
 <param name="input_mate_pairs">
 <collection type="paired">
 <element name="forward" value="bwa-mem-fastq1.fq" ftype="fastqsanger" />
 <element name="reverse" value="bwa-mem-fastq2.fq" ftype="fastqsanger" />
 </collection>
 <output_collection name="trimmed_reads_unpaired_collection" type="paired">
 <element name="forward" file="paired_collection_example_unpair1_results3.fastqsanger" ftype="fastqsanger"/>
 <element name="reverse" file="paired_collection_example_unpair2_results3.fastqsanger" ftype="fastqsanger"/>
 </output_collection>
 </test>
-<test>
+<test expect_num_outputs="7">
 <param name="input_mate_pairs">
 <collection type="paired">
 <element name="forward" value="bwa-mem-fastq1.fq.gz" ftype="fastqsanger.gz" />
 <element name="reverse" value="bwa-mem-fastq2.fq.gz" ftype="fastqsanger.gz" />
 </collection>
 <output_collection name="trimmed_reads_unpaired_collection" type="paired">
 <element name="forward" file="paired_collection_example_unpair1_results3.fastq.gz" ftype="fastqsanger.gz" decompress="true" />
 <element name="reverse" file="paired_collection_example_unpair2_results3.fastq.gz" ftype="fastqsanger.gz" decompress="true" />
 </output_collection>
 </test>
+<!--Test hard-trim option-->
+<test expect_num_outputs="2">
+<param name="input_mate1" value="bwa-mem-fastq1.fq" ftype="fastqsanger" />
+<param name="input_mate2" value="bwa-mem-fastq2.fq" ftype="fastqsanger" />
+<param name="sPaired" value="paired" />
+<conditional name="trimming">
+<param name="settingsType" value="custom" />
+<param name="hardtrim3" value="20"/>
+</conditional>
+<output name="hardtrimmed_reads_pair1" file="paired_hardtrimmed3_pair1_.fastqsanger" ftype="fastqsanger"/>
+<output name="hardtrimmed_reads_pair1" file="paired_hardtrimmed3_pair2_.fastqsanger" ftype="fastqsanger"/>
+</test>
+<test expect_num_outputs="2">
+<param name="input_mate1" value="bwa-mem-fastq1.fq" ftype="fastqsanger" />
+<param name="input_mate2" value="bwa-mem-fastq2.fq" ftype="fastqsanger" />
+<param name="sPaired" value="paired" />
+<conditional name="trimming">
+<param name="settingsType" value="custom" />
+<param name="hardtrim5" value="20"/>
+</conditional>
+<output name="hardtrimmed_reads_pair1" file="paired_hardtrimmed5_pair1_.fastqsanger" ftype="fastqsanger"/>
+<output name="hardtrimmed_reads_pair1" file="paired_hardtrimmed5_pair2_.fastqsanger" ftype="fastqsanger"/>
+</test>
+<!--Test mouse epigenetic clock option-->
+<test expect_num_outputs="2">
+<param name="input_mate1" value="bwa-mem-fastq1.fq" ftype="fastqsanger" />
+<param name="input_mate2" value="bwa-mem-fastq2.fq" ftype="fastqsanger" />
+<param name="sPaired" value="paired" />
+<conditional name="trimming">
+<param name="settingsType" value="custom" />
+<param name="clock" value="true"/>
+</conditional>
+<output name="mec_reads_pair1" file="mec_reads_pair1.fastqsanger" ftype="fastqsanger"/>
+<output name="mec_reads_pair2" file="mec_reads_pair2.fastqsanger" ftype="fastqsanger"/>
+</test>
+<!--Test polyA option-->
+<test expect_num_outputs="2">
+<param name="input_mate1" value="bwa-mem-fastq1.fq" ftype="fastqsanger" />
+<param name="input_mate2" value="bwa-mem-fastq2.fq" ftype="fastqsanger" />
+<param name="sPaired" value="paired" />
+<conditional name="trimming">
+<param name="settingsType" value="custom" />
+<param name="polyA" value="true"/>
+</conditional>
+<output name="trimmed_reads_pair1" file="trimmed_polyA_reads_pair1.fastqsanger" ftype="fastqsanger"/>
+<output name="trimmed_reads_pair1" file="trimmed_polyA_reads_pair2.fastqsanger" ftype="fastqsanger"/>
+</test>
 </tests>
 <help><![CDATA[
 **What it does**
 `Trim Galore!`_ is a wrapper script to automate quality and adapter trimming as well as quality control, with some added functionality to remove biased methylation positions for RRBS sequence files (for directional, non-directional (or paired-end) sequencing). It's main features are:
 * For any kind of FASTQ file other than MspI-digested RRBS, Trim Galore! can perform single-pass adapter and quality trimming
 * The Phred quality of basecalls and the stringency for adapter removal can be specified individually
 * Trim Galore! can remove sequences if they become too short during the trimming process. For paired-end files Trim Galore! removes entire sequence pairs if one (or both) of the two reads became shorter than the set length cutoff. Reads of a read-pair that are longer than a given threshold but for which the partner read has become too short can optionally be written out to single-end files. This ensures that the information of a read pair is not lost entirely if only one read is of good quality
 * Trim Galore! can trim paired-end files by 1 additional bp from the 3' end of all reads to avoid problems with invalid alignments with Bowtie 1
-.. _Trim Galore!: http://www.bioinformatics.babraham.ac.uk/projects/trim_galore/
 It is developed by Felix Krueger at the Babraham Institute.
 ----
 * **Screen quality-trimmed sequences for 'CAA' or 'CGA' at the start of the read and, if found, removes the first two basepairs**
 | Selecting this option for non-directional RRBS libraries will screen quality-trimmed sequences for 'CAA' or 'CGA' at the start of the read and, if found, removes the first two basepairs. Like with the option '--rrbs' this avoids using cytosine positions that were filled-in during the end-repair step. '--non_directional' requires '--rrbs' to be specified as well.
 |
 | *option --non_directional*
+----
+**Trim specific seetings**
+* **Hard-trimming mode**
+| Instead of performing adapter-/quality trimming, this option will simply hard-trim sequences to N bp at the 5' or 3' -end.
+|
+| *options -hardtrim5 and -hardtrim3*
+* **Mouse Epigenetic Clock mode**
+| In this mode, reads are trimmed in a specific way that is currently used for the Mouse Epigenetic Clock (see here: `Multi-tissue DNA methylation age predictor in mouse`_). Following this, Trim Galore will exit.
+|
+| In it's current implementation, the dual-UMI RRBS reads come in the following format:
+|
+| Read 1  5' UUUUUUUU CAGTA FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF TACTG UUUUUUUU 3'
+| Read 2  3' UUUUUUUU GTCAT FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF ATGAC UUUUUUUU 5'
+|
+| Where UUUUUUUU is a random 8-mer unique molecular identifier (UMI), CAGTA is a constant region, and FFFFFFF... is the actual RRBS-Fragment to be sequenced. The UMIs for Read 1 (R1) and Read 2 (R2), as well as the fixed sequences (F1 or F2), are written into the read ID and removed from the actual sequence.
+|
+| *option --clock*
+* **PolyA mode**
+| This is a new, still experimental, trimming mode to identify and remove poly-A tails from sequences. When selected, Trim Galore attempts to identify from the first supplied sample whether sequences contain more often a stretch of either 'AAAAAAAAAA' or 'TTTTTTTTTT'. This determines if Read 1 of a paired-end end file, or single-end files, are trimmed for PolyA or PolyT. In case of paired-end sequencing, Read2 is trimmed for the complementary base from the start of the reads.
+|
+| PLEASE NOTE: The poly-A trimming mode expects that sequences were both adapter and quality trimmed before looking for Poly-A tails, and it is the user's responsibility to carry out an initial round of trimming.
+|
+| *option --polyA*
+* **Amplicon mode**
+| This is a special mode of operation for paired-end data, such as required for the IMPLICON method, where a UMI sequence is getting transferred from the start of Read 2 to the readID of both reads. Following this, Trim Galore will exit.
+|
+| In it's current implementation, the UMI carrying reads come in the following format:
+|
+| Read 1  5' FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF 3'
+| Read 2  3' UUUUUUUUFFFFFFFFFFFFFFFFFFFFFFFFFFFF 5'
+|
+| Where UUUUUUUU is a random 8-mer unique molecular identifier (UMI) and FFFFFFF... is the actual fragment to be sequenced.
+|
+| *option --amplicon*
+.. _Trim Galore!: http://www.bioinformatics.babraham.ac.uk/projects/trim_galore/
+.. _Multi-tissue DNA methylation age predictor in mouse: https://genomebiology.biomedcentral.com/articles/10.1186/s13059-017-1203-5
 ]]></help>
-<citations></citations>
+<expand macro="citations" />
 </tool>

Mercurial > repos > bgruening > trim_galore

comparison trim_galore.xml @ 16:cd7e644cae1d draft default tip