Updating schema and check_samplesheet.py

nrhorner · Aug 21, 2023 · fac25ea · fac25ea
1 parent 28fe888
commit fac25ea
Show file tree

Hide file tree

Showing 2 changed files with 135 additions and 8 deletions.
diff --git a/assets/schema_input.json b/assets/schema_input.json
@@ -7,15 +7,47 @@
     "items": {
         "type": "object",
         "properties": {
+            "patient": {
+                "type": "string",
+                "pattern": "^\\S+$",
+                "errorMessage": "Patient ID must be provided and cannot contain spaces",
+                "meta": ["patient"]
+            },
             "sample": {
                 "type": "string",
                 "pattern": "^\\S+$",
-                "errorMessage": "Sample name must be provided and cannot contain spaces"
+                "errorMessage": "Sample ID must be provided and cannot contain spaces",
+                "meta": ["sample"]
             },
-            "fastq_1": {
+            "status": {
+                "type": "integer",
+                "errorMessage": "Status can only be 0 (normal), 1 (tumor) or 2 (rna tumour). Defaults to 1, if none is supplied.",
+                "meta": ["status"],
+                "default": "1",
+                "minimum": 0,
+                "maximum": 2
+            },
+            "lane": {
                 "type": "string",
-                "pattern": "^\\S+\\.f(ast)?q\\.gz$",
-                "errorMessage": "FastQ file for reads 1 must be provided, cannot contain spaces and must have extension '.fq.gz' or '.fastq.gz'"
+                "pattern": "^\\S+$",
+                "unique": ["patient", "sample"],
+                "dependentRequired": ["fastq_1"],
+                "meta": ["lane"]
+            },
+            "fastq_1": {
+                "errorMessage": "FastQ file for reads 1 cannot contain spaces and must have extension '.fq.gz' or '.fastq.gz'",
+                "anyOf": [
+                    {
+                        "type": "string",
+                        "pattern": "^\\S+\\.f(ast)?q\\.gz$"
+                    },
+                    {
+                        "type": "string",
+                        "maxLength": 0
+                    }
+                ],
+                "format": "file-path",
+                "exists": true
             },
             "fastq_2": {
                 "errorMessage": "FastQ file for reads 2 cannot contain spaces and must have extension '.fq.gz' or '.fastq.gz'",
@@ -28,9 +60,104 @@
                         "type": "string",
                         "maxLength": 0
                     }
-                ]
+                ],
+                "format": "file-path",
+                "exists": true
+            },
+            "table": {
+                "errorMessage": "Recalibration table cannot contain spaces and must have extension '.table'",
+                "anyOf": [
+                    {
+                        "type": "string",
+                        "pattern": "^\\S+\\.table$"
+                    },
+                    {
+                        "type": "string",
+                        "maxLength": 0
+                    }
+                ],
+                "format": "file-path",
+                "exists": true
+            },
+            "cram": {
+                "errorMessage": "CRAM file cannot contain spaces and must have extension '.cram'",
+                "anyOf": [
+                    {
+                        "type": "string",
+                        "pattern": "^\\S+\\.cram$"
+                    },
+                    {
+                        "type": "string",
+                        "maxLength": 0
+                    }
+                ],
+                "format": "file-path",
+                "exists": true
+            },
+            "crai": {
+                "errorMessage": "CRAM index file cannot contain spaces and must have extension '.crai'",
+                "anyOf": [
+                    {
+                        "type": "string",
+                        "pattern": "^\\S+\\.crai$"
+                    },
+                    {
+                        "type": "string",
+                        "maxLength": 0
+                    }
+                ],
+                "format": "file-path",
+                "exists": true
+            },
+            "bam": {
+                "errorMessage": "BAM file cannot contain spaces and must have extension '.bam'",
+                "anyOf": [
+                    {
+                        "type": "string",
+                        "pattern": "^\\S+\\.bam$"
+                    },
+                    {
+                        "type": "string",
+                        "maxLength": 0
+                    }
+                ],
+                "format": "file-path",
+                "exists": true
+            },
+            "bai": {
+                "errorMessage": "BAM index file cannot contain spaces and must have extension '.bai'",
+                "anyOf": [
+                    {
+                        "type": "string",
+                        "pattern": "^\\S+\\.bai$"
+                    },
+                    {
+                        "type": "string",
+                        "maxLength": 0
+                    }
+                ],
+                "format": "file-path",
+                "exists": true
+            },
+            "vcf": {
+                "errorMessage": "VCF file for reads 1 cannot contain spaces and must have extension '.vcf' or '.vcf.gz'",
+                "anyOf": [
+                    {
+                        "type": "string",
+                        "pattern": "^\\S+\\.vcf(\\.gz)?$"
+                    },
+                    {
+                        "type": "string",
+                        "maxLength": 0
+                    }
+                ],
+                "format": "file-path",
+                "exists": true
+            },
+            "variantcaller": {
+                "type": "string"
             }
         },
-        "required": ["sample", "fastq_1"]
+        "required": ["patient", "sample"]
     }
-}
+}
diff --git a/bin/check_samplesheet.py b/bin/check_samplesheet.py
@@ -71,7 +71,7 @@ def validate_and_transform(self, row):
         Perform all validations on the given row and insert the read pairing status.
 
         Args:
-            row (dict): A mapping from column headers (keys) to elements of that row
+            row (dict): A bam_align from column headers (keys) to elements of that row
                 (values).
 
         """