add cram into inputs

ghga-de · Nov 12, 2024 · 2b698ef · 2b698ef
1 parent 601f6a5
commit 2b698ef
Show file tree

Hide file tree

Showing 5 changed files with 36 additions and 10 deletions.
diff --git a/.gitignore b/.gitignore
@@ -13,6 +13,8 @@ results/
 result/
 results2/
 results37/
+results_vep/
+results_annovar/
 test.xml
 test_output/
 tests/data/
@@ -21,3 +23,4 @@ testdata_hg37/
 .github/CODEOWNERS-tmp
 bin/vcfparser.pyc
 singularity/
+snvs_*
diff --git a/assets/schema_input.json b/assets/schema_input.json
@@ -14,15 +14,33 @@
       },
       "tumor": {
         "type": "string",
-        "pattern": "^\\S+\\.bam$",
-        "errorMessage": "BAM file for tumors must be provided'"
+        "pattern": "^\\S+\\.(bam|cram)$",
+        "errorMessage": "BAM or CRAM file for tumors must be provided'"
+      },
+      "tumor_index": {
+        "type": "string",
+        "pattern": "^\\S+\\.(bai|crai)$",
+        "errorMessage": "BAI OR crai file matching to BAM for tumors must be provided'"
       },
       "control": {
-        "errorMessage": "BAM file for as control matching to tumor, if there is",
+        "errorMessage": "BAM or CRAM file for as control matching to tumor, if there is",
+        "anyOf": [
+          {
+            "type": "string",
+            "pattern": "^\\S+\\.(bam|cram)$"
+          },
+          {
+            "type": "string",
+            "maxLength": 0
+          }
+        ]
+      },
+      "control_index": {
+        "errorMessage": "BAI or CRAI file matching to BAM for as control matching to tumor, if there is",
         "anyOf": [
           {
             "type": "string",
-            "pattern": "^\\S+\\.bam$"
+            "pattern": "^\\S+\\.(bai|crai)$"
           },
           {
             "type": "string",
@@ -31,6 +49,6 @@
         ]
       }
     },
-    "required": ["sample", "tumor"]
+    "required": ["sample", "tumor", "tumor_index"]
   }
 }
diff --git a/conf/modules.config b/conf/modules.config
@@ -166,7 +166,7 @@ process {
         ]
     }
     withName: 'ENSEMBLVEP_VEP' {
-        //ext.args         ='--everything --filter_common --per_gene --total_length --offline'
+        ext.args         ='--everything --filter_common --per_gene --total_length --offline'
         publishDir       = [
             [
                 mode: params.publish_dir_mode,

diff --git a/modules/local/grep_samplename.nf b/modules/local/grep_samplename.nf
@@ -4,7 +4,7 @@ process GREP_SAMPLENAME {
 
     conda (params.enable_conda ? "" : null)
     container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
-        'docker://kubran/odcf_mpileupsnvcalling:v0':'kubran/odcf_mpileupsnvcalling:v0' }"
+        'docker://kubran/samtools:v1.9':'kubran/samtools:v1.9' }"
 
     input:
     tuple val(meta), path(tumor), path(tumor_bai), path(control),  path(control_bai)
@@ -20,8 +20,8 @@ process GREP_SAMPLENAME {
     if (meta.iscontrol == '1')
     {
         """
-        controlname=`samtools view -H $meta.control_bam | grep '^@RG' | sed "s/.*SM:\\([^\\t]*\\).*/\\1/g" | uniq`
-        tumorname=`samtools view -H $meta.tumor_bam | grep '^@RG' | sed "s/.*SM:\\([^\\t]*\\).*/\\1/g" | uniq`
+        controlname=`samtools view -H $control | grep '^@RG' | sed "s/.*SM:\\([^\\t]*\\).*/\\1/g" | uniq`
+        tumorname=`samtools view -H $tumor | grep '^@RG' | sed "s/.*SM:\\([^\\t]*\\).*/\\1/g" | uniq`
 
         cat <<-END_VERSIONS > versions.yml
         "${task.process}":
@@ -33,7 +33,7 @@ process GREP_SAMPLENAME {
     else {
         """
         controlname='dummy'
-        tumorname=`samtools view -H $meta.tumor_bam | grep '^@RG' | sed "s/.*SM:\\([^\\t]*\\).*/\\1/g" | uniq`
+        tumorname=`samtools view -H $tumor | grep '^@RG' | sed "s/.*SM:\\([^\\t]*\\).*/\\1/g" | uniq`
         
         cat <<-END_VERSIONS > versions.yml
         "${task.process}":

diff --git a/nextflow.config b/nextflow.config
@@ -82,6 +82,11 @@ params {
     cytobandcol                = '"CYTOBAND"'              // default value
     geneannocols               = '"ANNOVAR_FUNCTION,GENE,EXONIC_CLASSIFICATION,ANNOVAR_TRANSCRIPTS"' // default value
 
+    // VEP
+    vep_cache_version          = null
+    vep_genome                 = null
+    vep_cache                  = null
+    download_cache             = true  // DO NOT Download annotation cache
 
     // Files //
     data_path                  = null          // If database files in the same place