D3b RADIANT Annotation

Workflow to add various transcript and site level annotations to a DNA germline VCF. The workflow has the following framework:

• Old annotations are stripped with bcftools
• VCF is annotated with bcftools
• VCF is normalized
• VCF is annotated with VEP
• VCF is annotated with Echtvar
• Slivar Compound Hets is run
• Exomiser is run

Disabling Tools

Any of the tools can be enabled/disabled using the following disable_(insert_step) parameters. The default workflow runs in the following configuration:

• disable_bcftools_strip = false
• disable_bcftools_norm = false
• disable_vep = false
• disable_gnomad_anno = false
• disable_compound_hets = true
• disable_exomiser = true

While any step can be disabled, it might have adverse affects on downstream tools. For example, if you disable VEP but turn on Slivar compound hets, the program can fail if your input VCF does not have any ANN or CSQ INFO fields. Validation safeguards only exist for each step because we want to allow for scenarios where partially annotated VCFs can enter at any stage. This allows for faster reannotation in the future.

Inputs

Required - Need to provide

As mentioned above requirements depend on what is enabled. At the minimum, three inputs are required.

• vcf: KF VCF to process
• vcf_index = Index of vcf
• output_basename: String prefix of file outputs. Technically optional, but strongly recommended. Default is TEST

BCFTOOLS STRIP ENABLED

If you enable BCFTOOLS STRIP, at least one of the following must be provided:

• rm_fields_csv
• annotate_vcf
• bcftools_strip_extra_args

BCFTOOLS NORM ENABLED

If you enable BCFTOOLS NORM, all of the following must be provided:

• fasta

VEP ENABLED

If you enable VEP, all of the following must be provided:

• assembly
• fasta
• vep_cache
• vep_cache_version
• vep_species

ECHTVAR ENABLED

If you enable ECHTVAR, all of the following must be provided:

• echtvar_zips

SLIVAR ENABLED

If you enable SLIVAR, all of the following must be provided:

• ped

As mentioned above, Slivar does require some annotations. For more information see our notes.

EXOMISER ENABLED

• analysis_file: YAML with analysis options. See https://exomiser.readthedocs.io/en/latest/advanced_analysis.html#analysis. Recommend default_exomiser_WGS_analysis.yml
• datadir_file: TAR GZ with referenece files. Example contents:

  data/
  data/remm/
  data/remm/ReMM.v0.4.hg38.tsv.gz.tbi
  data/remm/ReMM.v0.4.hg38.tsv.gz
  data/2406_hg38.sha256
  data/2406_hg38/
  data/2406_hg38/2406_hg38_transcripts_refseq.ser
  data/2406_hg38/2406_hg38_transcripts_ensembl.ser
  data/2406_hg38/2406_hg38_genome.mv.db
  data/2406_hg38/2406_hg38_variants.mv.db
  data/2406_hg38/2406_hg38_transcripts_ucsc.ser
  data/2406_hg38/2406_hg38_clinvar.mv.db
  data/cadd/
  data/cadd/1.7/
  data/cadd/1.7/whole_genome_SNVs.tsv.gz
  data/cadd/1.7/whole_genome_SNVs.tsv.gz.tbi
  data/cadd/1.7/gnomad.genomes.r4.0.indel.tsv.gz
  data/cadd/1.7/gnomad.genomes.r4.0.indel.tsv.gz.tbi
  data/2406_phenotype/
  data/2406_phenotype/2406_phenotype.mv.db
  data/2406_phenotype/rw_string_10.mv
  data/2406_phenotype/phenix/
  data/2406_phenotype/phenix/out/
  data/2406_phenotype/phenix/out/10.out
  data/2406_phenotype/phenix/out/9_symmetric.out
  data/2406_phenotype/phenix/out/14.out
  data/2406_phenotype/phenix/out/13_symmetric.out
  data/2406_phenotype/phenix/out/20.out
  data/2406_phenotype/phenix/out/19.out
  data/2406_phenotype/phenix/out/16_symmetric.out
  data/2406_phenotype/phenix/out/7.out
  data/2406_phenotype/phenix/out/3_symmetric.out
  data/2406_phenotype/phenix/out/8.out
  data/2406_phenotype/phenix/out/17_symmetric.out
  data/2406_phenotype/phenix/out/15_symmetric.out
  data/2406_phenotype/phenix/out/5_symmetric.out
  data/2406_phenotype/phenix/out/11.out
  data/2406_phenotype/phenix/out/14_symmetric.out
  data/2406_phenotype/phenix/out/8_symmetric.out
  data/2406_phenotype/phenix/out/10_symmetric.out
  data/2406_phenotype/phenix/out/1.out
  data/2406_phenotype/phenix/out/20_symmetric.out
  data/2406_phenotype/phenix/out/11_symmetric.out
  data/2406_phenotype/phenix/out/18_symmetric.out
  data/2406_phenotype/phenix/out/4.out
  data/2406_phenotype/phenix/out/19_symmetric.out
  data/2406_phenotype/phenix/out/2.out
  data/2406_phenotype/phenix/out/12.out
  data/2406_phenotype/phenix/out/15.out
  data/2406_phenotype/phenix/out/2_symmetric.out
  data/2406_phenotype/phenix/out/9.out
  data/2406_phenotype/phenix/out/17.out
  data/2406_phenotype/phenix/out/16.out
  data/2406_phenotype/phenix/out/6_symmetric.out
  data/2406_phenotype/phenix/out/13.out
  data/2406_phenotype/phenix/out/18.out
  data/2406_phenotype/phenix/out/3.out
  data/2406_phenotype/phenix/out/4_symmetric.out
  data/2406_phenotype/phenix/out/1_symmetric.out
  data/2406_phenotype/phenix/out/7_symmetric.out
  data/2406_phenotype/phenix/out/6.out
  data/2406_phenotype/phenix/out/5.out
  data/2406_phenotype/phenix/out/12_symmetric.out
  data/2406_phenotype/phenix/hp.obo
  data/2406_phenotype/phenix/ALL_SOURCES_ALL_FREQUENCIES_genes_to_phenotype.txt
  data/2406_phenotype/hp.obo
  

• pheno_file: YAML file with sample information
  • Individual example:

      ---
      id: FM_ZCF17CPC
      proband:
      subject:
          id: BS_4JPDCXVR
          sex: MALE
      phenotypicFeatures:
          - type:
              id: HP:0009733
              label: Glioma
      pedigree:
      persons:
          - individualId: BS_4JPDCXVR
          sex: MALE
          affectedStatus: AFFECTED
    
      metaData:
      resources:
          - id: hp
          name: human phenotype ontology
          url: http://purl.obolibrary.org/obo/hp.owl
          version: hp/releases/2019-11-08
          namespacePrefix: HP
          iriPrefix: 'http://purl.obolibrary.org/obo/HP_'
      phenopacketSchemaVersion: 2.0

  • Family example:

     ---
     id: FM_ASMHHE2N
     proband:
     subject:
         id: BS_KPAR4N94
         sex: MALE
     phenotypicFeatures:
         - type:
             id: HP:0009733
     pedigree:
     persons:
         - individualId: BS_KPAR4N94
         sex: MALE
         paternalId: BS_264QQD38
         maternalId: BS_EQ98CJJT
         affectedStatus: AFFECTED
         - individualId: BS_264QQD38
         sex: MALE
         affectedStatus: UNAFFECTED
         - individualId: BS_EQ98CJJT
         sex: FEMALE
         affectedStatus: UNAFFECTED
     metaData:
     resources:
         - id: hp
         name: human phenotype ontology
         url: http://purl.obolibrary.org/obo/hp.owl
         version: hp/releases/2019-11-08
         namespacePrefix: HP
         iriPrefix: 'http://purl.obolibrary.org/obo/HP_'
     phenopacketSchemaVersion: 2.0
    

Input Defaults and Run Configurations

All of the defaults can be found in params block of nextflow.config. Kids First specific run configurations can be found in conf/kids_first.config.

Kids First Run Configuration

Here's a snapshot of the run configuration for Kids First Germline data:

• bcftools strip uses the annotate module to:
  • remove existing VEP (CSQ) and gnomad 3.1.1 (gnomad_3_1_1*) annotations
  • add annotations from annotate_vcf
• bcftools norm: Uses norm to split (or join) mulit-allelics and normalize indels
• VEP uses cache 105 and meets the following specific criteria:

  --allele_number
  --allow_non_variant
  --assembly GRCh38
  --cache
  --cache_version 105
  --canonical
  --ccds
  --check_existing
  --dir_cache ./
  --domains
  --dont_skip
  --failed 1
  --format vcf
  --gene_phenotype
  --hgvs
  --hgvsg
  --mane
  --mane_select
  --merged
  --no_escape
  --no_stats
  --numbers
  --offline
  --pick_order rank,biotype,mane,canonical,appris,tsl,ccds,length,ensembl,refseq
  --polyphen b
  --protein
  --pubmed
  --regulatory
  --shift_hgvs 1
  --sift b
  --species homo_sapiens
  --symbol
  --total_length
  --tsl
  --uniprot
  --variant_class
  --vcf
  --xref_refseq
  

• echtvar: Used for fast gnomAD annotation from the echtvar_zips input (gnomad.v3.1.1.custom.echtvar.zip)
• slivar: Used for rare variant discovery, primarily in joint family VCFs using the expr with the slivar_zips input (topmed.hg38.dbsnp.151.zip) and compount-hets modules
• exomiser: "The Exomiser is a Java program that finds potential disease-causing variants from whole-exome or whole-genome sequencing data." See https://github.com/exomiser/Exomiser for more details