@prefix LoadListingRequirement: .
@prefix CommandLineBinding: .
@prefix xsd: .
@prefix DockerRequirement: .
@prefix Workflow: .
@prefix sld: .
@prefix rdfs: .
@prefix CommandOutputBinding: .
@prefix ns2: .
@prefix ns1: .
@prefix NetworkAccess: .
@prefix rdf: .
@prefix cwl: .
cwl:in ;
cwl:out ;
cwl:run .
rdfs:comment "Raw to adjusted gene expression ratio plots" ;
cwl:outputSource ;
sld:type cwl:File , sld:null .
rdfs:comment "Compressed folder" ;
cwl:outputBinding [ CommandOutputBinding:glob
"*" ] ;
sld:type cwl:File .
rdfs:comment "Expression threshold for displaying target genes on a plot (expression > threshold)" ;
sld:type xsd:float , sld:null .
rdfs:comment "Round adjusted counts to integers" ;
sld:type xsd:boolean , sld:null .
rdfs:comment "Raw to adjusted gene expression ratio plots\n" ;
cwl:outputBinding [ CommandOutputBinding:glob
"*raw_to_adjusted_gene_expression_ratio_plots.pdf" ] ;
sld:type cwl:File , sld:null .
rdfs:comment "FDR cutoff for expression ratio plots" ;
sld:type xsd:float , sld:null .
rdfs:comment "Contamination estimation plot" ;
cwl:outputSource ;
sld:type cwl:File .
rdfs:comment "Adjusted feature-barcode matrices in HDF5 format" ;
cwl:outputSource ;
sld:type cwl:File .
rdfs:comment "SoupX stdout log" ;
cwl:outputSource ;
sld:type cwl:File .
cwl:in , , , , , , ;
cwl:out , , , , , , , , ;
cwl:run .
rdfs:comment "Adjusted feature-barcode matrices in HDF5 format\n" ;
cwl:outputBinding [ CommandOutputBinding:glob
"*adjusted_counts.h5" ] ;
sld:type cwl:File .
cwl:outputBinding [ CommandOutputBinding:glob
"soupx_stdout.log" ] ;
sld:type cwl:File .
cwl:outputBinding [ CommandOutputBinding:glob
"soupx_stderr.log" ] ;
sld:type cwl:File .
rdfs:comment "Compressed folder with secondary analysis results" ;
sld:type cwl:File .
cwl:inputBinding [ CommandLineBinding:position
2 ] ;
cwl:inputBinding [ CommandLineBinding:position
2 ] ;
sld:type [ sld:items cwl:File ;
sld:type sld:array
] ;
sld:type [ sld:items cwl:File ;
sld:type sld:array
] .
rdfs:comment "Target genes list. Headerless text file with 1 gene per line" ;
sld:type cwl:File , sld:null .
rdfs:comment "Compressed folder with unfiltered feature-barcode matrices" ;
sld:type cwl:File .
rdfs:comment "Expression threshold for displaying target genes on a plot (expression > threshold)\nDefault: 0.0\n" ;
cwl:inputBinding [ CommandLineBinding:prefix
"--threshold" ] ;
sld:type xsd:float , sld:null .
rdfs:comment "Round adjusted counts to integers\nDefault: False\n" ;
cwl:inputBinding [ CommandLineBinding:prefix
"--round" ] ;
sld:type xsd:boolean , sld:null .
cwl:source .
rdfs:comment "FDR cutoff for expression ratio plots\nDefault: 0.05\n" ;
cwl:inputBinding [ CommandLineBinding:prefix
"--fdr" ] ;
sld:type xsd:float , sld:null .
cwl:source .
rdfs:comment "Output matrix format version. Corresponds to the latest Cell Ranger matrix format" ;
sld:type , sld:null .
rdfs:comment "SoupX stderr log" ;
cwl:outputSource ;
sld:type cwl:File .
rdfs:comment "Raw gene expression plots" ;
cwl:outputSource ;
sld:type cwl:File , sld:null .
a cwl:CommandLineTool ;
rdfs:comment "Compresses input directory to tar.gz\n" ;
ns2:original_cwlVersion "v1.0" ;
ns1:codeRepository "https://github.com/Barski-lab/workflows" ;
ns1:creator [ a ns1:Organization ;
ns1:department [ a ns1:Organization ;
ns1:department [ a ns1:Organization ;
ns1:legalName "Barski Research Lab" ;
ns1:member [ a ns1:Person ;
ns1:email "mailto:misha.kotliar@gmail.com" ;
ns1:name "Michael Kotliar" ;
ns1:sameAs
]
] ;
ns1:legalName "Allergy and Immunology"
] ;
ns1:legalName "Cincinnati Children's Hospital Medical Center" ;
ns1:location [ a ns1:PostalAddress ;
ns1:addressCountry "USA" ;
ns1:addressLocality "Cincinnati" ;
ns1:addressRegion "OH" ;
ns1:postalCode "45229" ;
ns1:streetAddress "3333 Burnet Ave" ;
ns1:telephone "+1(513)636-4200"
] ;
ns1:logo "https://www.cincinnatichildrens.org/-/media/cincinnati%20childrens/global%20shared/childrens-logo-new.png"
] ;
ns1:downloadUrl "https://raw.githubusercontent.com/Barski-lab/workflows/master/tools/tar-compress.cwl" ;
ns1:isPartOf [ a ns1:CreativeWork ;
ns1:name "Common Workflow Language" ;
ns1:url "http://commonwl.org/"
] ;
ns1:license "http://www.apache.org/licenses/LICENSE-2.0" ;
ns1:name "tar-compress" ;
cwl:arguments ( [ CommandLineBinding:prefix "-C" ;
cwl:valueFrom "$(inputs.folder_to_compress.path.split(\"/\").slice(0,-1).join(\"/\"))"
]
"-czvf"
[ cwl:valueFrom "$(inputs.folder_to_compress.basename + \".tar.gz\")" ]
"."
) ;
cwl:baseCommand ( "tar" ) ;
cwl:cwlVersion ;
cwl:hints [ a ;
NetworkAccess:networkAccess true
] ;
cwl:hints [ a cwl:DockerRequirement ;
DockerRequirement:dockerPull "biowardrobe2/scidap:v0.0.3"
] ;
cwl:hints [ a ;
LoadListingRequirement:loadListing
"deep_listing"
] ;
cwl:inputs ;
cwl:outputs ;
cwl:requirements [ a cwl:InlineJavascriptRequirement ] .
rdfs:comment "Path to the file with target genes (headerless, one gene per line)\n" ;
cwl:inputBinding [ CommandLineBinding:prefix
"--genes" ] ;
sld:type cwl:File , sld:null .
sld:symbols , ;
sld:type sld:enum .
rdfs:comment "Compressed folder with filtered feature-barcode matrices" ;
sld:type cwl:File .
cwl:in ;
cwl:out ;
cwl:run .
cwl:source , , .
sld:symbols , ;
sld:type sld:enum .
cwl:outputBinding [ CommandOutputBinding:glob
"counts" ] ;
cwl:outputBinding [ CommandOutputBinding:glob
"counts" ] ;
sld:type cwl:Directory .
rdfs:comment "Compressed folder with adjusted feature-barcode matrices in MEX format" ;
cwl:outputSource ;
sld:type cwl:File .
rdfs:comment "Raw gene expression to pure soup ratio plots" ;
cwl:outputSource ;
sld:type cwl:File , sld:null .
cwl:source .
rdfs:comment "Adjusted gene expression plots\n" ;
cwl:outputBinding [ CommandOutputBinding:glob
"*adjusted_gene_expression_plots.pdf" ] ;
sld:type cwl:File , sld:null .
rdfs:comment "Folder to compressed" ;
sld:type cwl:Directory .
rdfs:comment "Contamination estimation plot\n" ;
cwl:outputBinding [ CommandOutputBinding:glob
"*contamination_estimation_plot.pdf" ] ;
sld:type cwl:File .
cwl:source .
rdfs:comment "Path to the output folder produced by 'cellranger count' command\n" ;
cwl:inputBinding [ CommandLineBinding:prefix
"--counts" ] ;
sld:type cwl:Directory .
rdfs:comment "Raw gene expression to pure soup ratio plots\n" ;
cwl:outputBinding [ CommandOutputBinding:glob
"*raw_gene_expression_to_pure_soup_ratio_plots.pdf" ] ;
sld:type cwl:File , sld:null .
a cwl:Workflow ;
rdfs:comment "Wrapped in a workflow SoupX tool for easy access to Cell Ranger pipeline compressed outputs.\n" ;
rdfs:label "SoupX (workflow) - an R package for the estimation and removal of cell free mRNA contamination" ;
ns2:original_cwlVersion "v1.0" ;
ns1:about "Wrapped in a workflow SoupX tool for easy access to Cell Ranger pipeline compressed outputs." ;
ns1:alternateName "SoupX (workflow) - an R package for the estimation and removal of cell free mRNA contamination" ;
ns1:codeRepository "https://github.com/Barski-lab/workflows" ;
ns1:creator [ a ns1:Organization ;
ns1:department [ a ns1:Organization ;
ns1:department [ a ns1:Organization ;
ns1:legalName "Barski Research Lab" ;
ns1:member [ a ns1:Person ;
ns1:email "mailto:misha.kotliar@gmail.com" ;
ns1:name "Michael Kotliar" ;
ns1:sameAs
]
] ;
ns1:legalName "Allergy and Immunology"
] ;
ns1:legalName "Cincinnati Children's Hospital Medical Center" ;
ns1:location [ a ns1:PostalAddress ;
ns1:addressCountry "USA" ;
ns1:addressLocality "Cincinnati" ;
ns1:addressRegion "OH" ;
ns1:postalCode "45229" ;
ns1:streetAddress "3333 Burnet Ave" ;
ns1:telephone "+1(513)636-4200"
] ;
ns1:logo "https://www.cincinnatichildrens.org/-/media/cincinnati%20childrens/global%20shared/childrens-logo-new.png"
] ;
ns1:downloadUrl "https://raw.githubusercontent.com/Barski-lab/workflows/master/tools/soupx-subworkflow.cwl" ;
ns1:isPartOf [ a ns1:CreativeWork ;
ns1:name "Common Workflow Language" ;
ns1:url "http://commonwl.org/"
] ;
ns1:license "http://www.apache.org/licenses/LICENSE-2.0" ;
ns1:name "SoupX (workflow) - an R package for the estimation and removal of cell free mRNA contamination" ;
Workflow:steps , , ;
cwl:cwlVersion ;
cwl:hints [ a ;
NetworkAccess:networkAccess true
] ;
cwl:hints [ a ;
LoadListingRequirement:loadListing
"deep_listing"
] ;
cwl:inputs , , , , , , , , ;
cwl:outputs , , , , , , , , ;
cwl:requirements [ a cwl:MultipleInputFeatureRequirement ] ;
cwl:requirements [ a cwl:InlineJavascriptRequirement ] .
cwl:default "set -- \"$0\" \"$@\"\nmkdir counts && cd counts\nfor i in \"$@\";\n do tar xvf $i;\ndone;\n" ;
cwl:inputBinding [ CommandLineBinding:position
1 ] ;
cwl:inputBinding [ CommandLineBinding:position
1 ] ;
sld:type xsd:string , sld:null .
a cwl:CommandLineTool ;
cwl:baseCommand ( "bash" "-c" ) ;
cwl:baseCommand ( "bash" "-c" ) ;
cwl:cwlVersion cwl:v1.0 ;
cwl:hints [ a cwl:DockerRequirement ;
DockerRequirement:dockerPull "scidap/scidap:v0.0.4"
] ;
cwl:hints [ a cwl:DockerRequirement ;
DockerRequirement:dockerPull "scidap/scidap:v0.0.4"
] ;
cwl:inputs , ;
cwl:outputs .
rdfs:comment "Folder with adjusted feature-barcode matrices in MEX format\n" ;
cwl:outputBinding [ CommandOutputBinding:glob
"*adjusted_counts" ] ;
sld:type cwl:Directory .
rdfs:comment "Output prefix\nDefault: soupx\n" ;
cwl:inputBinding [ CommandLineBinding:prefix
"--output" ] ;
sld:type xsd:string , sld:null .
cwl:source .
cwl:source .
rdfs:comment "Adjusted gene expression plots" ;
cwl:outputSource ;
sld:type cwl:File , sld:null .
cwl:source .
cwl:source .
rdfs:comment "Output matrix format version. Corresponds to the latest Cell Ranger matrix format\nDefault: 3\n" ;
cwl:inputBinding [ CommandLineBinding:prefix
"--format" ] ;
sld:type , sld:null .
a cwl:CommandLineTool ;
rdfs:comment "In droplet based, single cell RNA-seq experiments, there is always a certain amount of background\nmRNAs present in the dilution that gets distributed into the droplets with cells and sequenced\nalong with them. The net effect of this is to produce a background contamination that represents\nexpression not from the cell contained within a droplet, but the solution that contained the cells.\n\nThis collection of cell free mRNAs floating in the input solution (henceforth referred to as\n“the soup”) is created from cells in the input solution being lysed. Because of this, the soup looks\ndifferent for each input solution and strongly resembles the expression pattern obtained by summing\nall the individual cells.\n\nThe aim of this package is to provide a way to estimate the composition of this soup, what fraction\nof UMIs are derived from the soup in each droplet and produce a corrected count table with the soup\nbased expression removed.\n\nThe method to do this consists of three parts:\n\n- Calculate the profile of the soup.\n- Estimate the cell specific contamination fraction.\n- Infer a corrected expression matrix. \n" ;
rdfs:label "SoupX - an R package for the estimation and removal of cell free mRNA contamination" ;
ns2:original_cwlVersion "v1.0" ;
ns1:about "usage: run_soupx.R [-h] --counts COUNTS [--genes GENES]\n [--threshold THRESHOLD] [--fdr FDR]\n [--round] [--format {2,3}] [--output OUTPUT]\n\nRuns SoupX - an R package for the estimation and removal of cell free mRNA\ncontamination\n\noptional arguments:\n -h, --help show this help message and exit\n --counts COUNTS Path to the output folder produced by 'cellranger\n count' command\n --genes GENES Path to the file with target genes (headerless, one\n gene per line)\n --threshold THRESHOLD\n Expression threshold for displaying target genes on a\n plot (expression > threshold). Default: 0.0\n --fdr FDR FDR cutoff for expression ratio plots. Default: 0.05\n --round Round adjusted counts to integers\n --format {2,3} Output matrix format version. Default: 3 (corresponds\n to the latest Cell Ranger matrix format)\n --output OUTPUT Output prefix. Default: soupx" ;
ns1:alternateName "SoupX - an R package for the estimation and removal of cell free mRNA contamination" ;
ns1:codeRepository "https://github.com/Barski-lab/workflows" ;
ns1:creator [ a ns1:Organization ;
ns1:department [ a ns1:Organization ;
ns1:department [ a ns1:Organization ;
ns1:legalName "Barski Research Lab" ;
ns1:member [ a ns1:Person ;
ns1:email "mailto:misha.kotliar@gmail.com" ;
ns1:name "Michael Kotliar" ;
ns1:sameAs
]
] ;
ns1:legalName "Allergy and Immunology"
] ;
ns1:legalName "Cincinnati Children's Hospital Medical Center" ;
ns1:location [ a ns1:PostalAddress ;
ns1:addressCountry "USA" ;
ns1:addressLocality "Cincinnati" ;
ns1:addressRegion "OH" ;
ns1:postalCode "45229" ;
ns1:streetAddress "3333 Burnet Ave" ;
ns1:telephone "+1(513)636-4200"
] ;
ns1:logo "https://www.cincinnatichildrens.org/-/media/cincinnati%20childrens/global%20shared/childrens-logo-new.png"
] ;
ns1:downloadUrl "https://raw.githubusercontent.com/Barski-lab/workflows/master/tools/soupx.cwl" ;
ns1:isPartOf [ a ns1:CreativeWork ;
ns1:name "Common Workflow Language" ;
ns1:url "http://commonwl.org/"
] ;
ns1:license "http://www.apache.org/licenses/LICENSE-2.0" ;
ns1:name "SoupX - an R package for the estimation and removal of cell free mRNA contamination" ;
cwl:baseCommand ( "run_soupx.R" ) ;
cwl:cwlVersion ;
cwl:hints [ a ;
NetworkAccess:networkAccess true
] ;
cwl:hints [ a ;
LoadListingRequirement:loadListing
"deep_listing"
] ;
cwl:inputs , , , , , , ;
cwl:outputs , , , , , , , , ;
cwl:requirements [ a cwl:DockerRequirement ;
DockerRequirement:dockerPull "biowardrobe2/soupx:v0.0.1"
] ;
cwl:stderr "soupx_stderr.log" ;
cwl:stdout "soupx_stdout.log" .
rdfs:comment "Raw gene expression plots\n" ;
cwl:outputBinding [ CommandOutputBinding:glob
"*raw_gene_expression_plots.pdf" ] ;
sld:type cwl:File , sld:null .
rdfs:comment "Output prefix" ;
sld:type xsd:string , sld:null .