VCF
Definition
Variant Call Format, the standard output from variant callers (GATK, FreeBayes). Records each variant position with reference allele, alternative allele(s), genotype likelihoods, and quality filters. The forensic-STR tools parse VCF or BAM directly.
- Full name
- Variant Call Format
- Produced by
- Variant callers such as GATK and FreeBayes
- Records
- Position, reference/alternative alleles, genotype likelihoods, quality filters
- Forensic use
- Input format for downstream forensic STR/SNP allele-calling tools
Common questions
Why do forensic NGS pipelines need a step beyond the raw VCF output?+
General-purpose variant callers are built for clinical or research SNP and indel detection, not for the length-based repeat calling that forensic STR typing requires, so forensic-specific tools parse the VCF or the underlying BAM alignment to correctly count repeat units rather than relying on the caller's default variant model.
What quality information in a VCF matters most for a forensic conclusion?+
The genotype likelihood and quality filter fields indicate how confidently the caller distinguishes a true variant from sequencing noise or stutter artefact, which is central to deciding whether a called allele should be reported or treated as inconclusive in a mixed or low-template sample.
Related terms
- BAM File
- Binary Alignment/Map format, the compressed binary version of a SAM file. Stores aligned reads with their mapping positions, quality scores, and alignment...
- FASTQ
- The standard sequencing output file format containing four-line records per read: identifier, nucleotide sequence, separator, and Phred quality score string. All MPS...
- Isoallele
- Two STR alleles that share the same repeat count (and therefore appear as the same allele on CE) but differ by a...
- Phred Quality Score (Q)
- A logarithmic probability score for base-call accuracy. Q30 = 99.9% accuracy (1 in 1000 error probability). Q20 = 99% accuracy. Most forensic...
- Stutter (MPS)
- In MPS, stutter includes both length-based stutter (as in CE, an n-1 repeat artefact from PCR slippage) and sequence-based stutter (single-nucleotide errors...