somatic_SNVs_pipeline/tasks/bcftools.wdl

task bcftools {
  
  File ref_dir
  String fasta
  File vcf
  String basename = basename(vcf,".vcf")
  String docker
  String cluster_config
  String disk_size

  command <<<
    set -o pipefail
    set -e
    nt=$(nproc)
    
    # bcftools norm -m -both ${vcf} | bcftools norm -f ${ref_dir}/${fasta} -Ov -o ${basename}.norm.vcf
    # Split multiallelic sites
    bcftools norm -m -both ${vcf} -o ${basename}.norm.vcf
    bcftools view -O z -f 'PASS' ${basename}.norm.vcf -o ${basename}.PASS.vcf.gz
  >>>
  
  runtime {
    docker: docker
    cluster: cluster_config
    systemDisk: "cloud_ssd 40"
    dataDisk: "cloud_ssd " + disk_size + " /cromwell_root/"
  }
  
  output {
    File norm_vcf = "${basename}.PASS.vcf.gz"
  }
}