vcf

codescripts/contig.txt

+##contig=<ID=chr1,length=248956422>
+##contig=<ID=chr2,length=242193529>
+##contig=<ID=chr3,length=198295559>
+##contig=<ID=chr4,length=190214555>
+##contig=<ID=chr5,length=181538259>
+##contig=<ID=chr6,length=170805979>
+##contig=<ID=chr7,length=159345973>
+##contig=<ID=chr8,length=145138636>
+##contig=<ID=chr9,length=138394717>
+##contig=<ID=chr10,length=133797422>
+##contig=<ID=chr11,length=135086622>
+##contig=<ID=chr12,length=133275309>
+##contig=<ID=chr13,length=114364328>
+##contig=<ID=chr14,length=107043718>
+##contig=<ID=chr15,length=101991189>
+##contig=<ID=chr16,length=90338345>
+##contig=<ID=chr17,length=83257441>
+##contig=<ID=chr18,length=80373285>
+##contig=<ID=chr19,length=58617616>
+##contig=<ID=chr20,length=64444167>
+##contig=<ID=chr21,length=46709983>
+##contig=<ID=chr22,length=50818468>
+##contig=<ID=chrX,length=156040895>

inputs

   "{{ project_name }}.LCL5": "{{ LCL5 }}",
   "{{ project_name }}.project": "{{ project }}",
   "{{ project_name }}.SMALLcluster_config": "OnDemand bcs.ps.g.xlarge img-ubuntu-vpc",
-  "{{ project_name }}.BIGcluster_config": "OnDemand bcs.a2.7xlarge img-ubuntu-vpc",
+  "{{ project_name }}.BIGcluster_config": "OnDemand bcs.ps.g.2xlarge img-ubuntu-vpc",
   "{{ project_name }}.LCL6": "{{ LCL6 }}",
   "{{ project_name }}.sdf": "oss://pgx-reference-data/GRCh38.d1.vd1/GRCh38.d1.vd1.sdf/",
+  "{{ project_name }}.contig": "oss://pgx-result/renluyao/NCTR/contig.txt",
   "{{ project_name }}.MENDELIANdocker": "registry-vpc.cn-shanghai.aliyuncs.com/pgx-docker-registry/vbt:v1.1",
   "{{ project_name }}.LCL7": "{{ LCL7 }}",
   "{{ project_name }}.DIYdocker": "registry-vpc.cn-shanghai.aliyuncs.com/pgx-docker-registry/high_confidence_call_manuscript:v1.4",

tasks/benchmark.wdl

 task benchmark {
 	File vcf
 	File benchmarking_dir
+	File contig
 	File sdf
 	String project
 	String sample = basename(vcf,".raw.vcf.gz")
-	String fasta
 	String docker
 	String cluster_config
 	String disk_size
 		nt=$(nproc)
 		mkdir -p /cromwell_root/tmp
 		cp -r ${benchmarking_dir} /cromwell_root/tmp/
-        ls /cromwell_root/tmp/reference_datasets_v202103/ > files.txt
 
 		gunzip ${vcf} -c > unzip.vcf
+
+		cat unzip.vcf | grep '##' | grep -v 'contig' | cat - ${contig} > header
+		cat unzip.vcf | grep '#CHROM' > name
+		cat unzip.vcf | grep -v '#' > body
+		cat body | grep -w '^chr1\|^chr2\|^chr3\|^chr4\|^chr5\|^chr6\|^chr7\|^chr8\|^chr9\|^chr10\|^chr11\|^chr12\|^chr13\|^chr14\|^chr15\|^chr16\|^chr17\|^chr18\|^chr19\|^chr20\|^chr21\|^chr22\|^chrX' > body.filtered
+		cat header name body.filtered > filtered.vcf
+
  
-		/opt/rtg-tools/dist/rtg-tools-3.10.1-4d58ead/rtg bgzip unzip.vcf -c > ${project}.${sample}.rtg.vcf.gz
+		/opt/rtg-tools/dist/rtg-tools-3.10.1-4d58ead/rtg bgzip filtered.vcf -c > ${project}.${sample}.rtg.vcf.gz
 		/opt/rtg-tools/dist/rtg-tools-3.10.1-4d58ead/rtg index -f vcf ${project}.${sample}.rtg.vcf.gz
 		/opt/rtg-tools/dist/rtg-tools-3.10.1-4d58ead/rtg vcffilter -i ${project}.${sample}.rtg.vcf.gz -o ${sample}.rtg.SNV.vcf.gz --snps-only
 		/opt/rtg-tools/dist/rtg-tools-3.10.1-4d58ead/rtg vcffilter -i ${project}.${sample}.rtg.vcf.gz -o ${sample}.rtg.INDEL.vcf.gz --non-snps-only
 		File rtg_vcf_index = "${project}.${sample}.rtg.vcf.gz.tbi"
 		File SNV_result = "${project}.${sample}_SNV_precision_recall.txt"
 		File Indel_result = "${project}.${sample}_INDEL_precision_recall.txt"
-		File file_list = "files.txt"
 	}
 }

workflow.wdl

 	String fasta
 	File ref_dir
 	File sdf
+	File contig
 
 	File benchmarking_dir
 
 		vcf=LCL5,
 		benchmarking_dir=benchmarking_dir,
 		sdf=sdf,
+		contig=contig,
 		project=project,
-		fasta=fasta,
 		docker=BENCHMARKdocker,
 		cluster_config=BIGcluster_config,
 		disk_size=disk_size,
 		vcf=LCL6,
 		benchmarking_dir=benchmarking_dir,
 		sdf=sdf,
+		contig=contig,
 		project=project,
-		fasta=fasta,
 		docker=BENCHMARKdocker,
 		cluster_config=BIGcluster_config,
 		disk_size=disk_size,
 		vcf=LCL7,
 		benchmarking_dir=benchmarking_dir,
 		sdf=sdf,
+		contig=contig,
 		project=project,
-		fasta=fasta,
 		docker=BENCHMARKdocker,
 		cluster_config=BIGcluster_config,
 		disk_size=disk_size,
 		vcf=LCL8,
 		benchmarking_dir=benchmarking_dir,
 		sdf=sdf,
+		contig=contig,
 		project=project,
-		fasta=fasta,
 		docker=BENCHMARKdocker,
 		cluster_config=BIGcluster_config,
 		disk_size=disk_size,