The PCGR workflow accepts two types of input files:
- An unannotated, single-sample VCF file (>= v4.2) with called somatic variants (SNVs/InDels)
- A copy number segment file
PCGR can be run with either or both of the two input files present.
- We strongly recommend that the input VCF is compressed and indexed using bgzip and tabix
- If the input VCF contains multi-allelic sites, these will be subject to decomposition
- Variants used for reporting should be designated as ‘PASS’ in the VCF FILTER column
IMPORTANT NOTE 1: Considering the VCF output for the numerous somatic SNV/InDel callers that have been developed, we have a experienced a general lack of uniformity and robustness for the representation of somatic variant genotype data (e.g. variant allelic depths (tumor/normal), genotype quality etc.). Variant genotype data found as INFO tags in the input VCF can be specified as optional arguments to the PCGR workflow, which in turn can be used for interactive exploration in the tumor report.
IMPORTANT NOTE 2: PCGR generates a number of VCF INFO annotation tags that is appended to the query VCF. We will therefore encourage the users to submit query VCF files that have not been subject to annotations by other means, but rather a VCF file that comes directly from variant calling. If not, there are likely to be INFO tags in the query VCF file that coincide with those produced by PCGR.
The tab-separated values file with copy number aberrations MUST contain the following four columns:
- Chromosome
- Start
- End
- Segment_Mean
Here, Chromosome, Start, and End denote the chromosomal segment, and Segment_Mean denotes the log(2) ratio for a particular segment, which is a common output of somatic copy number alteration callers. Note that coordinates must be one-based (i.e. chromosomes start at 1, not 0). Below shows the initial part of a copy number segment file that is formatted correctly according to PCGR’s requirements:
Chromosome Start End Segment_Mean 1 3218329 3550598 0.0024 1 3552451 4593614 0.1995 1 4593663 6433129 -1.0277
The cancer genome sequencing report can be flexibly configured in a TOML-formatted configuration file. The default TOML configuration file, with descriptive comments wrt. usage are shown below:
# PCGR configuration options (TOML).
[tumor_only]
## If input VCF contains mix of germline/somatic (variants called with no matching control, i.e. tumor-only) set vcf_tumor_only to true
vcf_tumor_only = false
## if vcf_tumor_only = true, exclude variants (SNVs/InDels) with minor allele frequency above the following population-specific thresholds
## 1000 Genomes Project - WGS data
maf_onekg_eur = 0.002
maf_onekg_amr = 0.002
maf_onekg_afr = 0.002
maf_onekg_sas = 0.002
maf_onekg_eas = 0.002
maf_onekg_global = 0.002
## exclude variants with minor allele frequency above the following population-specific thresholds
## gnomAD - WES data
maf_gnomad_nfe = 0.002
maf_gnomad_amr = 0.002
maf_gnomad_afr = 0.002
maf_gnomad_sas = 0.002
maf_gnomad_eas = 0.002
maf_gnomad_fin = 0.002
maf_gnomad_oth = 0.002
maf_gnomad_global = 0.002
## Exclude variants occurring in PoN (panel of normals, if provided as VCF)
exclude_pon = true
## Exclude likely homozygous germline variants (100% allelic fraction for alternate allele in tumor, very unlikely somatic event)
exclude_likely_hom_germline = false
## Exclude likely heterozygous germline variants
## Must satisfy i) 40-60 % allelic fraction for alternate allele in tumor sample, ii) present in dbSNP + gnomAD, ii) not existing as somatic event in COSMIC/TCGA
## Note that the application of this filter may be suboptimal for very impure tumors or variants affected by CNAs etc (under these circumstances, the allelic fraction
## will be skewed (see e.g. discussion in PMID:29249243)
exclude_likely_het_germline = false
## Exclude variants found in dbSNP (only those that are NOT found in ClinVar(somatic origin)/DoCM/TCGA/COSMIC)
exclude_dbsnp_nonsomatic = false
## exclude all non-exonic variants
exclude_nonexonic = true
[allelic_support]
## Specify INFO tags in input VCF that denotes depth/allelic fraction in tumor and normal sample
## An additional tag that denotes call confidence (call_conf_tag) can also be specified, which will
## be used for exploration in the global variant browser. Note that 'tumor_dp_tag' must be of
## Type=Integer, and 'tumor_af_tag' must be of Type=Float (similarly for normal sample)
tumor_dp_tag = ""
tumor_af_tag = ""
normal_dp_tag = ""
normal_af_tag = ""
call_conf_tag = ""
## set thresholds for tumor/normal depth/allelic fraction, will be applied before report generation
## requires that 'tumor/normal_dp_tag' and 'tumor/normal_af_tag' are specified above
tumor_dp_min = 0
tumor_af_min = 0.0
normal_dp_min = 0
normal_af_max = 1.0
[mutational_burden]
## Calculate mutational burden (similar to Chalmers et al., Genome Med, 2017)
mutational_burden = true
## Size of coding target region in megabases (defaults to exome ~ 36 Mb)
## Note: this should ideally denote the callable target size (i.e. reflecting variable
## sequencing depth)
target_size_mb = 34.0
## set upper limits to tumor mutational burden tertiles (mutations/Mb)
tmb_low_limit = 5
tmb_intermediate_limit = 20
## tmb_high = tmb > tmb_intermediate_limit
[cna]
## log ratio thresholds for determination of copy number gains and homozygous deletions
logR_gain = 0.8
logR_homdel = -0.8
## percent overlap between copy number segment and gene transcripts (average) for reporting of gains/losses in tumor suppressor genes/oncogenes
cna_overlap_pct = 100
[msi]
## Predict microsatellite instability
msi = true
[mutational_signatures]
## Identify relative contribution of 30 known mutational signatures (COSMIC) through the deconstructSigs framework
mutsignatures = true
## deconstructSigs option: number of mutational signatures to limit the search to ('signatures.limit' in whichSignatures)
mutsignatures_signature_limit = 6
## deconstructSigs option: type of trimer count normalization for inference of known mutational signatures, see explanation at https://github.com/raerose01/deconstructSigs"
## options = 'default', 'exome', 'genome', 'exome2genome'
## NOTE: If your data (VCF) is from exome sequencing, 'default' or 'exome2genome' should be used. See https://github.com/raerose01/deconstructSigs/issues/2
mutsignatures_normalization = "exome2genome"
## Require a minimum number of mutations for signature estimation
mutsignatures_mutation_limit = 100
## deconstructSigs option: discard any signature contributions with a weight less than this amount
mutsignatures_cutoff = 0.06
[visual]
## Choose visual theme of report, any of: "default", "cerulean", "journal", "flatly", "readable", "spacelab", "united", "cosmo", "lumen", "paper", "sandstone", "simplex", or "yeti" (https://bootswatch.com/)
report_theme = "default"
[custom_tags]
## list VCF info tags that should be present in JSON and TSV output
## tags should be comma separated, i.e. custom_tags = "MUTECT2_FILTER,STRELKA_FILTER"
custom_tags = ""
[other]
## list/do not list noncoding variants
list_noncoding = true
## VEP/vcfanno processing options
n_vcfanno_proc = 4
n_vep_forks = 4
## Customise the order of criteria used to pick the primary transcript in VEP (see https://www.ensembl.org/info/docs/tools/vep/script/vep_options.html#opt_pick_order)
vep_pick_order = "canonical,appris,biotype,ccds,rank,tsl,length"
## omit intergenic variants during VEP processing
vep_skip_intergenic = false
## generate a MAF for input VCF using https://github.com/mskcc/vcf2maf
vcf2maf = true