nextflow_schema.json

{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "$id": "https://raw.githubusercontent.com/Flanders//nextflow_schema.json",
  "title": "Flanders pipeline parameters",
  "description": "Finemapping and colocalisation pipeline",
  "type": "object",
  "$defs": {
    "input_output_options": {
      "title": "Input/output options",
      "type": "object",
      "fa_icon": "fas fa-terminal",
      "description": "Define where the pipeline should find input data and save output data.",
      "required": ["outdir"],
      "properties": {
        "summarystats_input": {
          "type": "string",
          "description": "Input TSV file containing summary stats to be processed and corresponding configuration",
          "format": "file-path",
          "mimetype": "text/tsv",
          "schema": "assets/summarystats_input_schema.json",
          "fa_icon": "fas fa-file-csv"
        },
        "coloc_h5ad_input": {
          "type": "string",
          "description": "Input h5ad file(s) containing results from previous finemapping runs to be loaded",
          "format": "file-path"
        },
        "coloc_exclude_studies_table": {
          "type": "string",
          "description": "A tab-separated file with header study_id, phenotype_id containing a list of traits to exclude from colocalisation analysis",
          "format": "file-path",
          "mimetype": "text/tsv",
          "fa_icon": "fas fa-file-csv"
        },
        "coloc_guide_table": {
          "type": "string",
          "description": "A guide table for colocalisation. When provided will take precedence",
          "format": "file-path",
          "mimetype": "text/csv",
          "fa_icon": "fas fa-file-csv"
        },
        "outdir": {
          "type": "string",
          "format": "directory-path",
          "description": "The output directory where the results will be saved. You have to use absolute paths to storage on Cloud infrastructure.",
          "fa_icon": "fas fa-folder-open",
          "default": "flanders_output"
        }
      }
    },
    "finemapping": {
      "title": "Finemapping",
      "type": "object",
      "description": "",
      "default": "",
      "properties": {
        "finemap_id": {
          "type": "string",
          "description": "An ID for the finemapping results, will be used as prefix in the h5ad finemapping output",
          "default": "finemap_run"
        },
        "chromosomes": {
          "type": "string",
          "description": "A string setting chromosomes to include in munging and finemapping. Intervals string can be like 1-22 or 1,6,3 or 1-6,7-10",
          "default": "1-22"
        },
        "run_liftover": {
          "type": "boolean",
          "default": true,
          "description": "Set to true to liftover coordinates to GRCh38"
        },
        "skip_dentist": {
          "type": "boolean",
          "default": true,
          "hidden": true,
          "description": "Perform DENTIST cleanup before finemapping. Usually not needed"
        },
        "large_locus_size": {
          "type": "integer",
          "default": 2000000,
          "description": "Maximum size for a locus to enter finemapping and colocalization analysis"
        },
        "susie_max_iter": {
          "type": "integer",
          "default": 400,
          "description": "Maximum number of IBSS iterations to perform when running SUSIE, see https://stephenslab.github.io/susieR/reference/susie.html"
        },
        "susie_qc_cs_bf_thr": {
          "type": "number",
          "default": 3,
          "description": "Credible set BF threshold for credible sets QC"
        },
        "susie_qc_pval_thr": {
          "type": "number",
          "default": 1,
          "description": "Minimum SNP p-value threshold for credible sets QC"
        },
        "susie_qc_mean_r2_thr": {
          "type": "number",
          "default": 0,
          "description": "Credible set purity mean r2 threshold for credible sets QC"
        },
        "susie_qc_min_r2_thr": {
          "type": "number",
          "default": 0,
          "description": "Credible set for purity minimum r2 threshold for credible sets QC"
        },
        "publish_susie": {
          "type": "boolean",
          "default": false,
          "hidden": true,
          "description": "Whether to publish the susie finemap .rds intermediate files"
        }
      },
      "required": [
        "finemap_id",
        "large_locus_size",
        "susie_max_iter",
        "susie_qc_cs_bf_thr",
        "susie_qc_pval_thr",
        "susie_qc_mean_r2_thr",
        "susie_qc_min_r2_thr"
      ], 
      "fa_icon": "fas fa-bezier-curve"
    },
    "colocalization": {
      "title": "Colocalization",
      "type": "object",
      "description": "",
      "default": "",
      "properties": {
        "run_colocalization": {
          "type": "boolean",
          "default": true,
          "description": "Set to true to run colocalization after finemapping"
        },
        "coloc_id": {
          "type": "string",
          "description": "An ID for the colocalization results, will be used as prefix in colocalization outputs",
          "default": "coloc_run"
        },
        "coloc_skip_previous_studies": {
          "type": "boolean",
          "default": false,
          "description": "In case a previous h5ad input is given, do not evaluate colocalizations that involves only previous loci"
        },
        "coloc_batch_size": {
          "type": "integer",
          "default": 5000,
          "hidden": true,
          "description": "Number of loci to be colocalized per batch"
        },
        "pph3_threshold": {
          "type": "number",
          "default": 0.75,
          "description": "The PP.H3 ABF threshold value to save results into a separate H3 table"
        },
        "pph4_threshold": {
          "type": "number",
          "default": 0.75,
          "description": "The PP.H4 ABF threshold value to save results into a separate H4 table"
        }
      },
      "required": [
        "coloc_id",
        "coloc_batch_size",
        "pph3_threshold",
        "pph4_threshold"
      ],
      "fa_icon": "fas fa-map-marker-alt"
    },
    "advanced_parameters": {
      "title": "Advanced parameters",
      "type": "object",
      "description": "",
      "default": "",
      "properties": {
        "publish_dir_mode": {
          "type": "string",
          "default": "copy",
          "hidden": true,
          "description": "Define the publish mode for result files generated by the pipeline"
        },
        "is_test_profile": {
          "type": "boolean",
          "default": false,
          "hidden": true,
          "description": "Set to true when running test profile"
        }
      },
      "fa_icon": "fas fa-tools",
      "required": ["publish_dir_mode"]
    },
    "institutional_config_options": {
      "title": "Institutional config options",
      "type": "object",
      "fa_icon": "fas fa-university",
      "description": "Parameters used to describe centralised config profiles. These should not be edited.",
      "help_text": "These parameters define centralized configuration profiles that appear in the Nextflow log when you run a pipeline. You typically don't need to modify these values.",
      "properties": {
        "custom_config_version": {
          "type": "string",
          "description": "Git commit id for Institutional configs.",
          "default": "master",
          "hidden": true,
          "fa_icon": "fas fa-users-cog"
        },
        "custom_config_base": {
          "type": "string",
          "description": "Base directory for Institutional configs.",
          "default": "https://raw.githubusercontent.com/nf-core/configs/master",
          "hidden": true,
          "help_text": "When running offline, Nextflow cannot retrieve institutional configuration files from the internet. If needed, download these files from the repository and specify their location with this parameter.",
          "fa_icon": "fas fa-users-cog"
        },
        "config_profile_name": {
          "type": "string",
          "description": "Institutional config name.",
          "hidden": true,
          "fa_icon": "fas fa-users-cog"
        },
        "config_profile_description": {
          "type": "string",
          "description": "Institutional config description.",
          "hidden": true,
          "fa_icon": "fas fa-users-cog"
        },
        "config_profile_contact": {
          "type": "string",
          "description": "Institutional config contact information.",
          "hidden": true,
          "fa_icon": "fas fa-users-cog"
        },
        "config_profile_url": {
          "type": "string",
          "description": "Institutional config URL link.",
          "hidden": true,
          "fa_icon": "fas fa-users-cog"
        },
        "max_cpus": {
          "type": "integer",
          "description": "Maximum number of CPUs available in the system",
          "hidden": true,
          "fa_icon": "fas fa-users-cog"
        },
        "max_memory": {
          "type": "string",
          "description": "Maximum amount of memory available in the system",
          "hidden": true,
          "fa_icon": "fas fa-users-cog"
        },
        "max_time": {
          "type": "string",
          "description": "Maximum wall time available for a job in the system",
          "hidden": true,
          "fa_icon": "fas fa-users-cog"
        },
        "igenomes_base": {
          "type": "string",
          "description": "Base directory for iGenomes reference data.",
          "format": "directory-path",
          "hidden": true,
          "fa_icon": "fas fa-folder-open"
        }
      }
    }
  },
  "allOf": [
    {
      "$ref": "#/$defs/input_output_options"
    },
    {
      "$ref": "#/$defs/finemapping"
    },
    {
      "$ref": "#/$defs/colocalization"
    },
    {
      "$ref": "#/$defs/advanced_parameters"
    },
    {
      "$ref": "#/$defs/institutional_config_options"
    }
  ]
}