Alignment_Based_CNV_eval

.dockstore.yml

@@ -117,5 +117,10 @@ workflows:
  - name: CNV_BLAST
    subclass: WDL
    primaryDescriptorPath: /CNV_BLAST/cnv_blast.wdl
+   testParameterFiles:
+    - /CNV_BLAST/cnv_blast.inputs.json
+ - name: AlignmentBasedCNVEvaluation
+   subclass: WDL
+   primaryDescriptorPath: /AlignmentBasedCNVEvaluation/AlignmentBasedCNVEvaluation.wdl
    testParameterFiles:
     - /CNV_BLAST/cnv_blast.inputs.json

AlignmentBasedCNVEvaluation/AlignmentBasedCNVEvaluation.inputs.json

@@ -0,0 +1,12 @@
+{
+  "AlignmentBasedCNVEval.reference_fasta": "File",
+  "AlignmentBasedCNVEval.T2T_last_database": "File",
+  "AlignmentBasedCNVEval.extract_cnv.mem_size": "Int (optional, default = 4)",
+  "AlignmentBasedCNVEval.extract_cnv.cpu_num": "Int (optional, default = 1)",
+  "AlignmentBasedCNVEval.cnv_vcf": "File",
+  "AlignmentBasedCNVEval.analysis_docker": "String",
+  "AlignmentBasedCNVEval.reference_last_database": "File",
+  "AlignmentBasedCNVEval.last_docker": "String",
+  "AlignmentBasedCNVEval.extract_cnv.disk_size": "Int (optional, default = 10)"
+}
+

AlignmentBasedCNVEvaluation/AlignmentBasedCNVEvaluation.wdl

@@ -0,0 +1,60 @@
+version 1.0
+
+    workflow AlignmentBasedCNVEval {
+        input {
+            File cnv_vcf
+            File reference_fasta
+            File reference_last_database
+            File T2T_last_database
+            String last_docker
+            String analysis_docker
+        }
+        call extract_cnv {
+            input:
+                cnv_vcf = cnv_vcf
+        }
+    }
+
+    task extract_cnv {
+            input {
+                File cnv_vcf
+                Int cpu_num = 1
+                Int mem_size = 4
+                Int disk_size = 10
+            }
+            command <<<
+                set -e
+
+                filename=$(basename ~{cnv_vcf})
+                # Unzip the input file if it is gzipped
+                if test "${filename##*.}" = "gz"; then
+                  gunzip -c ~{cnv_vcf} > temp.vcf
+                else
+                  cp ~{cnv_vcf} temp.vcf
+                fi
+
+                # Extract the PASS events from the DRAGEN CNV VCF and save them to a file
+                cat temp.vcf | grep -v '#' | awk '$7 == "PASS" {print}' | awk '{print $3}' | awk 'BEGIN {FS=":"} {print $1":"$2"\t"$3":"$4}' > pass_cnv_event.txt
+
+                # Split CNV intervals to individual files
+                split -l 100 -d --additional-suffix=_interval.txt pass_cnv_event.txt pass_cnv_chunk_
+
+                # Print the number of GAINS and LOSSES that DRAGEN found
+                cat temp.vcf | grep -v '#' | awk '$7 == "PASS" {print}' | awk '{print $3}' | awk 'BEGIN {FS=":"} {print $2}' | sort | uniq -c | sort
+            >>>
+            output {
+                File uncompressed_cnv_vcf = "temp.vcf"
+                File cnv_events_file = "pass_cnv_event.txt"
+                Array[File] cnv_intervals_chucks = glob("pass_cnv_chunk_*")
+
+            }
+            runtime {
+                docker: "us.gcr.io/broad-dsde-methods/liquidbiopsy:0.0.3.7"
+                bootDiskSizeGb: 12
+                cpu: cpu_num
+                memory: mem_size + " GB"
+                disks: "local-disk " + disk_size + " HDD"
+                preemptible: 2
+                #maxRetries: 3
+            }
+        }