schema_input.json

{
    "$schema": "http://json-schema.org/draft-07/schema",
    "$id": "https://raw.githubusercontent.com/sanger-tol/treeval/master/assets/schema_input.json",
    "title": "sanger-tol/treeval pipeline - params.input schema",
    "description": "Schema for the file provided with params.input",
    "type": "array",
    "items": {
        "type": "object",
        "properties": {
            "assembly": {
                "type": "object",
                "properties": {
                    "sample_id": {
                        "type": "string",
                        "errorMessage": "Sample name must be provided and cannot contain spaces"
                    },
                    "latin_name": {
                        "type": "string",
                        "errorMessage": "The scientific name for the assembly"
                    },
                    "classT": {
                        "type": "string",
                        "errorMessage": "The Clade of the assembly. Used as the syntenic group and to complete the gene_alignment data dir."
                    },
                    "TicketType": {
                        "type": "string",
                        "errorMessage": "Not currently in use. Single word description of associated project."
                    }
                }
            },
            "reference_file": {
                "type": "string",
                "pattern": "^\\S+\\.f(ast)a$",
                "errorMessage": "Assembly input file, decompressed"
            },
            "assem_reads": {
                "type": "object",
                "properties": {
                    "pacbio": {
                        "type": "string",
                        "errorMessage": "Path to folder containing fasta.gz files"
                    },
                    "hic": {
                        "type": "string",
                        "errorMessage": "Path to folder containing .cram and .crai files"
                    },
                    "supplementary": {
                        "type": "string",
                        "errorMessage": "Not currently in use. Placeholder for future use"
                    }
                }
            },
            "alignment": {
                "type": "object",
                "properties": {
                    "data_dir": {
                        "type": "string",
                        "errorMessage": "Gene Alignment data directory"
                    },
                    "common_name": {
                        "type": "string",
                        "errorMessage": "Not currently in use. Common identifier for group (adding bee, wasp, ant as sub division for clade)"
                    },
                    "geneset": {
                        "type": "string",
                        "errorMessage": "A csv list of organisms to run against."
                    }
                }
            },
            "self_comp": {
                "type": "object",
                "properties": {
                    "motif_len": {
                        "type": "integer",
                        "errorMessage": "Length of motif to be used in self comp search"
                    },
                    "mummer_chunk": {
                        "type": "integer",
                        "errorMessage": "Size of chunks to be used my Mummer"
                    }
                }
            },
            "synteny": {
                "type": "object",
                "properties": {
                    "synteny_genome_path": {
                        "type": "string",
                        "errorMessage": "Syntenic Genome Directory Path"
                    }
                }
            },
            "outdir": {
                "type": "string",
                "errorMessage": "Out directory path, can be changed via cli"
            },
            "intron": {
                "type": "object",
                "properties": {
                    "size": {
                        "type": "string",
                        "errorMessage": "Base pair size of introns, defaults to 50k"
                    }
                }
            },
            "telomere": {
                "type": "object",
                "properties": {
                    "teloseq": {
                        "type": "string",
                        "errorMessage": "Expected telomeric motif"
                    }
                }
            },
            "busco": {
                "type": "object",
                "properties": {
                    "lineage_path": {
                        "type": "string",
                        "errorMessage": "Path to directory containing lineages folder"
                    },
                    "lineage": {
                        "type": "string",
                        "errorMessage": "busco lineage to run"
                    }
                }
            }
        },
        "required": [
            "busco",
            "telomere",
            "intron",
            "synteny",
            "self_comp",
            "alignment",
            "assem_reads",
            "reference_file",
            "assembly"
        ]
    }
}