nextflow_schema.json

{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "$id": "https://raw.githubusercontent.com/break-through-cancer/tcrtoolkit-pipeline/main/nextflow_schema.json",
  "title": "tcrtoolkit pipeline parameters",
  "description": "BTC TCR Toolkit pipeline",
  "type": "object",

  "$defs": {
    "input_output_options": {
      "title": "Input/output options",
      "type": "object",
      "fa_icon": "fas fa-terminal",
      "description": "Define where the pipeline should find input data and save output data.",
      "required": ["samplesheet", "outdir"],
      "properties": {
        "samplesheet": {
          "type": "string",
          "format": "file-path",
          "pattern": ".*.csv$",
          "description": "Path to the samplesheet describing input AIRR data.",
          "help_text": "A CSV of samples and metadata for this TCR analysis.",
          "fa_icon": "fas fa-file-csv"
        },
        "outdir": {
          "type": "string",
          "format": "directory-path",
          "default": "out",
          "description": "Output directory where results will be saved.",
          "fa_icon": "fas fa-folder-open"
        }
      }
    },

    "resource_options": {
      "title": "Max resource options",
      "type": "object",
      "fa_icon": "fab fa-acquisitions-incorporated",
      "description": "Set the top limit for requested resources for any single job.",
      "properties": {
        "max_cpus": {
          "type": "integer",
          "default": 192,
          "description": "Maximum CPUs that can be requested by any process.",
          "fa_icon": "fas fa-microchip"
        },
        "max_memory": {
          "type": "string",
          "default": "768.GB",
          "pattern": "^\\d+(?:\\.\\d+|\\.)?\\s*(?:[KMGT]?B|[KMGT])$",
          "description": "Maximum memory for any process.",
          "fa_icon": "fas fa-memory"
        },
        "max_time": {
          "type": "string",
          "default": "48.h",
          "pattern": "^\\d+(?:\\.\\d+|\\.)?\\s*(s|m|h|d)$",
          "description": "Maximum walltime for any job.",
          "fa_icon": "far fa-clock"
        }
      }
    },

    "workflow_options": {
      "title": "Workflow parameters",
      "type": "object",
      "fa_icon": "fas fa-project-diagram",
      "description": "General pipeline workflow settings.",
      "properties": {
        "workflow_level": {
          "type": "string",
          "default": "sample,compare",
          "description": "Comma-separated workflow stages.",
          "pattern": "^(sample|patient|compare|convert)(,(sample|patient|compare|convert))*$"
        },
        "project_name": {
          "type": "string",
          "description": "Name of this analysis project."
        },
        "publish_dir_mode": {
          "type": "string",
          "default": "copy",
          "enum": ["copy", "move", "link", "symlink"],
          "description": "Method used by `publishDir` to save outputs."
        }
      }
    },

    "airr_options": {
      "title": "AIRR data options",
      "type": "object",
      "fa_icon": "fas fa-dna",
      "description": "Parameters related to AIRR format and schema references.",
      "properties": {
        "input_format": {
          "type": "string",
          "default": "airr",
          "enum": ["airr", "adaptive", "cellranger"],
          "description": "Input data format."
        },
        "airr_schema": {
          "type": "string",
          "description": "Path to AIRR rearrangement schema JSON."
        },
        "imgt_lookup": {
          "type": "string",
          "description": "Path to imgt lookup table."
        },
        "sample_stats_template": {
          "type": "string",
          "description": "Path to sample notebook template."
        },
        "compare_stats_template": {
          "type": "string",
          "description": "Path to compare notebook template."
        }
      }
    },

    "plotting_options": {
      "title": "Plotting and metadata options",
      "type": "object",
      "fa_icon": "fas fa-chart-bar",
      "description": "Parameters for plotting and metadata columns.",
      "properties": {
        "samplechart_x_col": { "type": "string", "default": "timepoint" },
        "samplechart_color_col": { "type": "string", "default": "origin" },
        "vgene_subject_col": { "type": "string", "default": "subject_id" },
        "vgene_x_cols": { "type": "string", "default": "origin,timepoint" }
      }
    },

    "giana_options": {
      "title": "GIANA clustering options",
      "type": "object",
      "fa_icon": "fas fa-brain",
      "properties": {
        "threshold": { "type": "number", "default": 7.0 },
        "threshold_score": { "type": "number", "default": 3.6 },
        "threshold_vgene": { "type": "number", "default": 3.7 }
      }
    },

    "gliph2_options": {
      "title": "GLIPH2 clustering options",
      "type": "object",
      "fa_icon": "fas fa-code-branch",
      "properties": {
        "gliph2_report_template": { "type": "string" },
        "ref_files": { "type": "string" },
        "local_min_pvalue": { "type": "string", "default": "0.001" },
        "p_depth": { "type": "string", "default": "1000" },
        "global_convergence_cutoff": { "type": "string", "default": "1" },
        "simulation_depth": { "type": "string", "default": "1000" },
        "kmer_min_depth": { "type": "string", "default": "3" },
        "local_min_OVE": { "type": "string", "default": "c(1000, 100, 10)" },
        "algorithm": { "type": "string", "default": "GLIPH2" },
        "all_aa_interchangeable": { "type": "string", "default": "1" }
      }
    },

    "tcrdist3_options": {
      "title": "TCRdist3 distance options",
      "type": "object",
      "fa_icon": "fas fa-ruler-combined",
      "properties": {
        "matrix_sparsity": {
          "type": "string",
          "default": "sparse",
          "enum": ["sparse", "full"]
        },
        "distance_metric": { "type": "string", "default": "tcrdist" },
        "db_path": { "type": "string" }
      }
    }
  },

  "allOf": [
    { "$ref": "#/$defs/input_output_options" },
    { "$ref": "#/$defs/resource_options" },
    { "$ref": "#/$defs/workflow_options" },
    { "$ref": "#/$defs/airr_options" },
    { "$ref": "#/$defs/plotting_options" },
    { "$ref": "#/$defs/giana_options" },
    { "$ref": "#/$defs/gliph2_options" },
    { "$ref": "#/$defs/tcrdist3_options" }
  ]
}