Watershed pipeline for SNVs (WDL implementation)

Description of inputs:

• File cadd_anno_header: File containing VCF header lines describing the annotations.
  • See the VCF 4.2 specification for full details.
  • For example, one line could be: ##INFO=<ID=SIFTval,Number=1,Type=String,Description="SIFT score">
  • It is best to set Type=String to be robust to missing values which are coded in unpredictable formats.
• File cadd_cache: CADD annotations in tabular format.
  • For example, this file of CADD v1.6 annotations which can be downloaded from https://cadd.gs.washington.edu.
• File cadd_cache_idx: Tabix index (.tbi) file for cadd_cache.
• File cadd_cols2keep: File indicating which columns of the . See bcftools annotate -C documentation for full details, but briefly:
  • Columns must be listed in order they appear in cadd_cache.
  • Columns representing chromosome, position, reference and alternate alleles must be labelled CHROM,POS,REF,ALT.
  • Columns to drop are listed as -. Columns to keep are given a name.
• File chr_rename_file: A file with two columns of chromosome codes: one of the chromosome names in your vcfs, and the other with chromosomes named as 1,2,...22,X.
  • This is used to make vcfs which have the chromosome naming scheme chr1,chr2... etc. compatible with the cadd_cache`.
• File chr_unrename_file: Similar to chr_rename_file, but maps the chromosome codes back to how they were before.
• File gerp_bw: BigWig (.bw) file of GERP scores downloadable here (used by VEP's loftee plugin).
• File human_ancestor_seq: Human ancestor sequence file downloadable here (used by VEP's loftee plugin).
• File phylocsf_db: SQL database of PhyloCSF metrics downloadable here (used by VEP's loftee plugin).
• File phylop100_bw: BigWig (.bw) file of phyloP100way scores, downloadable from UCSC here.
  • These scores represent the degree to which variants are conserved in a collection of 100 non-human vertebrate species. For more information, see this page of the UCSC Genome Browser site.
• Array[File] vcfs: VCF (or BCF) file(s) to be annotated.
  • The files must contain INFO/AC and INFO/AN fields at minimum.
• File vep_cache: v115 of the cache file for Ensembl's Variant Effect Predictor (VEP), (downloadable here).
• (Optional) File filter_regions: File of regions to filter the vcfs by, one region per line.
  • See bcftools documentation about the -R option for full details.
• (Optional) File filter_samples: File of sample ids to filter the vcfs by, one id per line.
  • See bcftools documentation about the -S option for full details.
• (Optional) Int n_cpu: Number of cores to allocate. More cores will make the workflow finish more quickly, but also cost slightly more.
  • For example, a run that took 3hr:45min and $1.75 on 8 cores, took 1hr:30min and $2.73 on 32 cores.