{"@id":"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl","retrievedFrom":{"repoUrl":"https://github.com/EBI-Metagenomics/workflow-is-cwl.git","branch":"26dad276bac124f89086268bcbca962a5c0caca6","path":"workflows/TranscriptomeAssembly-wf.paired-end.cwl","packedId":null,"type":"GITHUB","rawUrl":"https://raw.githubusercontent.com/EBI-Metagenomics/workflow-is-cwl/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl","url":"https://github.com/EBI-Metagenomics/workflow-is-cwl/blob/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl"},"retrievedOn":"2023-01-12T19:42:37.316+00:00","lastCommit":"26dad276bac124f89086268bcbca962a5c0caca6","label":"Transcriptome assembly workflow (paired-end version)","inputs":{"end_mode":{"label":"read -end mode format","doc":"Read -end mode format to be specify to Trimmomatic\n","type":"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/tools/Trimmomatic/trimmomatic-end_mode.yaml#end_mode"},"read_files":{"label":"FASTQ read file(s)","doc":"FASTQ file of reverse reads in Paired End mode\n","type":"File[] [FASTQ]","format":"http://edamontology.org/format_1930"},"trinity_cpu":{"label":"number of CPUs allocated","doc":"number of CPUs to use, default: 2\n","type":"Integer (Optional)"},"forward_reads":{"label":"Paired-end read file 1","doc":"Read file 1 in FASTQ format \n","type":"File [FASTQ]","format":"http://edamontology.org/format_1930"},"reverse_reads":{"label":"Paired-end read file 2","doc":"Read file 2 in FASTQ format\n","type":"File [FASTQ]","format":"http://edamontology.org/format_1930"},"trinity_max_mem":{"label":"maximum memory allocated to Trinity","doc":"Suggested max memory to use by Trinity where limiting can be enabled. (jellyfish, sorting, etc) provided in Gb of RAM, ie. --max_memory 10G\n","type":"String"},"trinity_seq_type":{"label":"read file(s) format","doc":"type of reads: (fa or fq)\n","type":"String"},"trimmomatic_phred":{"label":"quality score format","doc":"Either PHRED \\\"33\\\" or \\\"64\\\" specifies the base quality encoding. Default: 64\n","type":"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/tools/Trimmomatic/trimmomatic-phred.yaml#phred"},"trinity_ss_lib_type":{"label":"Strand-specific RNA-Seq read orientation","doc":"Strand-specific RNA-Seq read orientation. if paired: RF or FR, if single: F or R. (dUTP method = RF). See web documentation\n","type":"String"},"trimmomatic_slidingWindow":{"label":"read filtering sliding window","doc":"Perform a sliding window trimming, cutting once the average quality within the window falls below a threshold. By considering multiple bases, a single poor quality base will not cause the removal of high quality data later in the read. <windowSize> specifies the number of bases to average across <requiredQuality> specifies the average quality required\n","type":"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/tools/Trimmomatic/trimmomatic-sliding_window.yaml#slidingWindow"}},"outputs":{"raw_qc_report":{"type":"File[]"},"raw_html_report":{"type":"File[] [HTML]","format":"http://edamontology.org/format_2331"},"assembled_contigs":{"type":"File [FASTA]","format":"http://edamontology.org/format_1929"},"filtered_qc_report":{"type":"File[]"},"assembly_output_dir":{"type":"Directory"},"filtered_html_report":{"type":"File[] [HTML]","format":"http://edamontology.org/format_2331"},"forward_reads_paired":{"type":"File [FASTQ]","format":"http://edamontology.org/format_1930"},"reverse_reads_paired":{"type":"File [FASTQ]","format":"http://edamontology.org/format_1930"},"transrate_output_dir":{"type":"Directory"},"trimmomatic_log_file":{"type":"File (Optional)"},"forward_reads_unpaired":{"type":"File (Optional) [FASTQ]","format":"http://edamontology.org/format_1930"},"reverse_reads_unpaired":{"type":"File (Optional) [FASTQ]","format":"http://edamontology.org/format_1930"}},"steps":{"filter_reads":{"label":"Trimmomatic - A flexible read trimming tool for Illumina NGS data","doc":"Trimmomatic is a fast, multithreaded command line tool that can be used to trim and crop Illumina (FASTQ) data as well as to remove adapters. These adapters can pose a real problem depending on the library preparation and downstream application. There are two major modes of the program: Paired end mode and Single end mode. The paired end mode will maintain correspondence of read pairs and also use the additional information contained in paired reads to better find adapter or PCR primer fragments introduced by the library preparation process. Trimmomatic works with FASTQ files (using phred + 33 or phred + 64 quality scores, depending on the Illumina pipeline used).\n","run":"../tools/Trimmomatic/Trimmomatic-v0.36.cwl","runType":"COMMANDLINETOOL","sources":{"filter_reads":{"defaultVal":"3"},"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl#filter_reads/phred":{"sourceIDs":["trimmomatic_phred"]},"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl#filter_reads/minlen":{"defaultVal":"100"},"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl#filter_reads/reads1":{"sourceIDs":["forward_reads"]},"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl#filter_reads/reads2":{"sourceIDs":["reverse_reads"]},"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl#filter_reads/end_mode":{"sourceIDs":["end_mode"]},"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl#filter_reads/trailing":{"defaultVal":"3"},"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl#filter_reads/slidingwindow":{"sourceIDs":["trimmomatic_slidingWindow"]}}},"run_assembly":{"label":"Trinity assembles transcript sequences from Illumina RNA-Seq data.","doc":"Trinity, developed at the Broad Institute and the Hebrew University of Jerusalem,  represents a novel method for the efficient and robust de novo reconstruction  of transcriptomes from RNA-seq data.  Trinity combines three independent software modules: Inchworm, Chrysalis, and  Butterfly, applied sequentially to process large volumes of RNA-seq reads.\nDocumentation at https://github.com/trinityrnaseq/trinityrnaseq/wiki\n","run":"../tools/Trinity/Trinity-V2.6.5.paired-end.cwl","runType":"COMMANDLINETOOL","sources":{"run_assembly":{"sourceIDs":["trinity_seq_type"]},"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl#run_assembly/left_reads":{"sourceIDs":["filter_reads"]},"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl#run_assembly/right_reads":{"sourceIDs":["filter_reads"]},"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl#run_assembly/trinity_cpu":{"sourceIDs":["trinity_cpu"]},"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl#run_assembly/trinity_max_mem":{"sourceIDs":["trinity_max_mem"]},"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl#run_assembly/trinity_ss_lib_type":{"sourceIDs":["trinity_ss_lib_type"]}}},"evaluate_contigs":{"label":"Transrate - A de-novo transcriptome assembly evaluation facility.","doc":"Analyse a de-novo transcriptome assembly using three kinds of metrics: 1. sequence based (if --assembly is given) 2. read mapping based (if --left and --right are given) 3. reference based (if --reference is given)\nDocumentation at http://hibberdlab.com/transrate\n      \n","run":"../tools/Transrate/Transrate-V1.0.3.cwl","runType":"COMMANDLINETOOL","sources":{"evaluate_contigs":{"sourceIDs":["filter_reads"]},"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl#evaluate_contigs/in_fasta":{"sourceIDs":["run_assembly"]},"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl#evaluate_contigs/left_fastq":{"sourceIDs":["filter_reads"]}}},"generate_raw_stats":{"label":"FastQC - A high throughtput sequence analyses QC.","doc":"FastQC reads a set of sequence files and produces from each one a quality control report consisting of a number of different modules, each one of which will help to identify a different potential type of problem in your data.\nIf no files to process are specified on the command line then the program will start as an interactive graphical application.  If files are provided on the command line then the program will run with no user interaction required.  In this mode it is suitable for inclusion into a standardised analysis pipeline.\n  \nPlease visit https://www.bioinformatics.babraham.ac.uk/projects/fastqc/ for full documentation.\n  \n","run":"../tools/FastQC/FastQC-v0.11.7.cwl","runType":"COMMANDLINETOOL","sources":{"generate_raw_stats":{"sourceIDs":["read_files"]}}},"generate_filtered_stats":{"label":"FastQC - A high throughtput sequence analyses QC.","doc":"FastQC reads a set of sequence files and produces from each one a quality control report consisting of a number of different modules, each one of which will help to identify a different potential type of problem in your data.\nIf no files to process are specified on the command line then the program will start as an interactive graphical application.  If files are provided on the command line then the program will run with no user interaction required.  In this mode it is suitable for inclusion into a standardised analysis pipeline.\n  \nPlease visit https://www.bioinformatics.babraham.ac.uk/projects/fastqc/ for full documentation.\n  \n","run":"../tools/FastQC/FastQC-v0.11.7.cwl","runType":"COMMANDLINETOOL","sources":{"generate_filtered_stats":{"sourceIDs":["filter_reads"]}}}},"cwltoolVersion":"3.1.20221201130942","visualisationDot":"digraph workflow {\n  graph [\n    bgcolor = \"#eeeeee\"\n    color = \"black\"\n    fontsize = \"10\"\n    labeljust = \"left\"\n    clusterrank = \"local\"\n    ranksep = \"0.22\"\n    nodesep = \"0.05\"\n  ]\n  node [\n    fontname = \"Helvetica\"\n    fontsize = \"10\"\n    fontcolor = \"black\"\n    shape = \"record\"\n    height = \"0\"\n    width = \"0\"\n    color = \"black\"\n    fillcolor = \"lightgoldenrodyellow\"\n    style = \"filled\"\n  ];\n  edge [\n    fontname=\"Helvetica\"\n    fontsize=\"8\"\n    fontcolor=\"black\"\n    color=\"black\"\n    arrowsize=\"0.7\"\n  ];\n  subgraph cluster_inputs {\n    rank = \"same\";\n    style = \"dashed\";\n    label = \"Workflow Inputs\";\n    \"reverse_reads\" [fillcolor=\"#94DDF4\",label=\"Paired-end read file 2\"];\n    \"trinity_seq_type\" [fillcolor=\"#94DDF4\",label=\"read file(s) format\"];\n    \"trimmomatic_slidingWindow\" [fillcolor=\"#94DDF4\",label=\"read filtering sliding window\"];\n    \"end_mode\" [fillcolor=\"#94DDF4\",label=\"read -end mode format\"];\n    \"read_files\" [fillcolor=\"#94DDF4\",label=\"FASTQ read file(s)\"];\n    \"trinity_cpu\" [fillcolor=\"#94DDF4\",label=\"number of CPUs allocated\"];\n    \"trinity_max_mem\" [fillcolor=\"#94DDF4\",label=\"maximum memory allocated to Trinity\"];\n    \"trinity_ss_lib_type\" [fillcolor=\"#94DDF4\",label=\"Strand-specific RNA-Seq read orientation\"];\n    \"trimmomatic_phred\" [fillcolor=\"#94DDF4\",label=\"quality score format\"];\n    \"forward_reads\" [fillcolor=\"#94DDF4\",label=\"Paired-end read file 1\"];\n  }\n  subgraph cluster_outputs {\n    rank = \"same\";\n    style = \"dashed\";\n    labelloc = \"b\";\n    label = \"Workflow Outputs\";\n    \"filtered_html_report\" [fillcolor=\"#94DDF4\",label=\"filtered_html_report\"];\n    \"reverse_reads_paired\" [fillcolor=\"#94DDF4\",label=\"reverse_reads_paired\"];\n    \"assembled_contigs\" [fillcolor=\"#94DDF4\",label=\"assembled_contigs\"];\n    \"transrate_output_dir\" [fillcolor=\"#94DDF4\",label=\"transrate_output_dir\"];\n    \"forward_reads_paired\" [fillcolor=\"#94DDF4\",label=\"forward_reads_paired\"];\n    \"reverse_reads_unpaired\" [fillcolor=\"#94DDF4\",label=\"reverse_reads_unpaired\"];\n    \"raw_html_report\" [fillcolor=\"#94DDF4\",label=\"raw_html_report\"];\n    \"filtered_qc_report\" [fillcolor=\"#94DDF4\",label=\"filtered_qc_report\"];\n    \"assembly_output_dir\" [fillcolor=\"#94DDF4\",label=\"assembly_output_dir\"];\n    \"raw_qc_report\" [fillcolor=\"#94DDF4\",label=\"raw_qc_report\"];\n    \"trimmomatic_log_file\" [fillcolor=\"#94DDF4\",label=\"trimmomatic_log_file\"];\n    \"forward_reads_unpaired\" [fillcolor=\"#94DDF4\",label=\"forward_reads_unpaired\"];\n  }\n  \"generate_filtered_stats\" [label=\"FastQC - A high throughtput sequence analyses QC.\"];\n  \"evaluate_contigs\" [label=\"Transrate - A de-novo transcriptome assembly evaluation facility.\"];\n  \"filter_reads\" [label=\"Trimmomatic - A flexible read trimming tool for Illumina NGS data\"];\n  \"generate_raw_stats\" [label=\"FastQC - A high throughtput sequence analyses QC.\"];\n  \"run_assembly\" [label=\"Trinity assembles transcript sequences from Illumina RNA-Seq data.\"];\n  \"filter_reads\" -> \"generate_filtered_stats\" [label=\"in_fastq\"];\n  \"filter_reads\" -> \"evaluate_contigs\" [label=\"right_fastq\"];\n  \"filter_reads\" -> \"evaluate_contigs\" [label=\"left_fastq\"];\n  \"run_assembly\" -> \"evaluate_contigs\" [label=\"in_fasta\"];\n  \"default1\" -> \"filter_reads\" [label=\"leading\"];\n  \"default1\" [label=\"3\", fillcolor=\"#D5AEFC\"];\n  \"trimmomatic_slidingWindow\" -> \"filter_reads\" [label=\"slidingwindow\"];\n  \"end_mode\" -> \"filter_reads\" [label=\"end_mode\"];\n  \"reverse_reads\" -> \"filter_reads\" [label=\"reads2\"];\n  \"default2\" -> \"filter_reads\" [label=\"trailing\"];\n  \"default2\" [label=\"3\", fillcolor=\"#D5AEFC\"];\n  \"default3\" -> \"filter_reads\" [label=\"minlen\"];\n  \"default3\" [label=\"100\", fillcolor=\"#D5AEFC\"];\n  \"trimmomatic_phred\" -> \"filter_reads\" [label=\"phred\"];\n  \"forward_reads\" -> \"filter_reads\" [label=\"reads1\"];\n  \"read_files\" -> \"generate_raw_stats\" [label=\"in_fastq\"];\n  \"trinity_seq_type\" -> \"run_assembly\" [label=\"trinity_seq_type\"];\n  \"filter_reads\" -> \"run_assembly\" [label=\"right_reads\"];\n  \"trinity_max_mem\" -> \"run_assembly\" [label=\"trinity_max_mem\"];\n  \"trinity_ss_lib_type\" -> \"run_assembly\" [label=\"trinity_ss_lib_type\"];\n  \"filter_reads\" -> \"run_assembly\" [label=\"left_reads\"];\n  \"trinity_cpu\" -> \"run_assembly\" [label=\"trinity_cpu\"];\n  \"generate_filtered_stats\" -> \"filtered_html_report\";\n  \"filter_reads\" -> \"reverse_reads_paired\";\n  \"run_assembly\" -> \"assembled_contigs\";\n  \"evaluate_contigs\" -> \"transrate_output_dir\";\n  \"filter_reads\" -> \"forward_reads_paired\";\n  \"filter_reads\" -> \"reverse_reads_unpaired\";\n  \"generate_raw_stats\" -> \"raw_html_report\";\n  \"generate_filtered_stats\" -> \"filtered_qc_report\";\n  \"run_assembly\" -> \"assembly_output_dir\";\n  \"generate_raw_stats\" -> \"raw_qc_report\";\n  \"filter_reads\" -> \"trimmomatic_log_file\";\n  \"filter_reads\" -> \"forward_reads_unpaired\";\n}\n","licenseLink":"https://spdx.org/licenses/Apache-2.0","packed":false,"permalink":"https://w3id.org/cwl/view/git/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl","visualisationXdot":"/graph/xdot/github.com/EBI-Metagenomics/workflow-is-cwl/blob/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl","roBundle":"/robundle/github.com/EBI-Metagenomics/workflow-is-cwl/blob/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl","licenseName":"Apache License 2.0","visualisationPng":"/graph/png/github.com/EBI-Metagenomics/workflow-is-cwl/blob/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl","visualisationSvg":"/graph/svg/github.com/EBI-Metagenomics/workflow-is-cwl/blob/26dad276bac124f89086268bcbca962a5c0caca6/workflows/TranscriptomeAssembly-wf.paired-end.cwl"}