callers.yaml

# All paths are relative to the directory that Snakemake is executed in
# Note: this file is written in the YAML syntax (https://learnxinyminutes.com/docs/yaml/)

# CALLER SPECIFIC PARAMETERS
# These parameters are used by the pipeline when it executes each of the callers.
# All caller specific parameters are optional except where otherwise noted as
# "required" in the comments below.
# You can specify:
#   1) cols - any other columns (besides CHROM, POS, REF, and ALT) to extract from the VCFs output by each caller
#      (note that this attribute will be ignored if your caller script outputs a TSV instead of a VCF)
#   2) params - any extra parameters that should be passed to the caller script
#   3) ext - whether your caller outputs a VCF (assumed if not specified) or a TSV
#   4) na - replacement values to use (instead of the default value of 0) when filling NA's
#      (note that this is ignored if 'keep_na' is set to true in config.yaml or prepare.yaml)
# Each key should be the <caller_id> and each value can be a dictionary containing any of the attributes listed above
# "cols" should be a dictionary of lists for each category of columns in the VCF (where categories can be 'info', 'format', and 'other' for all others)
# "params" can be a single string or a list of strings
# "na" should be a dictionary containing as keys each column name and as values the replacement value
# Note that the trained classification models we provide in the example data
# only work with these settings. So if you change any of the "cols" or "na"
# values, you will also need to create new trained classification models.
gatk-snp:
  cols:
    other: [QUAL]
    info: [QD,FS,MQ,AC,ExcessHet]
    format: [DP,GQ]

varscan-snp:
  cols:
    info: [ADP]
    format: [SDP,DP,RD,AD,PVAL,ABQ,GQ,RBQ,RDF,RDR,ADF,ADR]
  na:
    PVAL: 1.01E0

vardict-snp:
  cols:
    other: [QUAL]
    info: [TYPE,DP,VD,AF,QSTD,MQ,SN,HIAF,ADJAF,NM,SVTYPE,SVLEN,DUPRATE]
  na:
    SVTYPE: NC

pg-snp:
  # path to a bgzip, indexed VCF containing variants called in the platinum genomes project
  # required!
  params: /iblm/netapp/data1/external/PlatinumGenomes/2017-1.0/hg19/hybrid/hg19.hybrid.no_chr.vcf.gz

gatk-indel:
  cols:
    info: [QD,FS,MQ,AC,ExcessHet]
    format: [DP,GQ]

varscan-indel:
  cols:
    info: [ADP]
    format: [SDP,DP,RD,AD,PVAL,ABQ,GQ,RBQ,RDF,RDR,ADF,ADR]
  na:
    PVAL: 1.01E0

vardict-indel:
  cols:
    other: [QUAL]
    info: [TYPE,DP,VD,AF,QSTD,MQ,SN,HIAF,ADJAF,NM,SVTYPE,SVLEN,DUPRATE]
  na:
    SVTYPE: NC

delly:
  cols:
    format: [RC,GQ]

pindel:
  cols:
    info: [HOMLEN,SVLEN,SVTYPE]
    format: [PL,RD]
  na:
    SVTYPE: NC

illumina-manta:
  cols:
    other: [QUAL]
    info: [SVTYPE,SVLEN,HOMLEN,BND_DEPTH,MATE_BND_DEPTH]
    format: [GQ]
  na:
    SVTYPE: NC

illumina-strelka:
  cols:
    other: [QUAL]
    info: [REFREP,IDREP,MQ]
    format: [GQX,DP,DPF,MIN_DP,DPI,SB]
  params:
    # 1) path to strelka config file
    # required!
    - configs/configureStrelkaGermlineWorkflow.py.ini
    # 2) path to the directory in which strelka is stored or "" if strelka is installed in your current conda env (or if using --use-conda)
    # required!
    - ""

breakca:
  # specify that breakca creates a tsv instead of a vcf
  # required!
  ext: tsv
  params:
    # 1) the (relative) path to the breakCA scripts directory
    # required!
    - breakCA
    # 2) path to the directory in which the Rscript executable is held or "" if R is installed in your current conda env (or if using --use-conda)
    # required!
    - ""

pg-indel:
  # path to a bgzip, indexed VCF containing variants called in the platinum genomes project
  # required!
  params: /iblm/netapp/data1/external/PlatinumGenomes/2017-1.0/hg19/hybrid/hg19.hybrid.no_chr.vcf.gz

# Custom parameters can also be passed to special scripts that perform pre-caller steps:
illumina:
  params:
    # 1) path to manta config file
    # required!
    - configs/configManta.py.ini
    # 2) path to the directory in which manta is stored or "" if manta is installed in your current conda env (or if using --use-conda)
    # required!
    - ""

pg:
  # path to a bgzip, indexed VCF containing variants called in the platinum genomes project
  # required!
  params: /iblm/netapp/data1/external/PlatinumGenomes/2017-1.0/hg19/hybrid/hg19.hybrid.no_chr.vcf.gz