Understanding VCF Files

intermediate

VCF Field Structure

Anatomy of a VCF Data Line

📍CHROMchr1 ... chrX, chrY, chrM

🔢POS1-based position on chromosome

🏷IDrsID from dbSNP or . if novel

🔄REF/ALTReference and alternate alleles

⭐QUALPhred confidence variant exists

🔍FILTERPASS or filter name (LowQual)

📊INFODP, AF, gnomAD_AF, CADD...

🧬FORMAT/GT0/1 het, 1/1 hom-alt, 0/0 ref

VCF Structure

code

## Meta-information lines
##fileformat=VCFv4.2
##reference=file:///data/hg38.fa
##contig=<ID=chr1,length=248956422>
##INFO=<ID=DP,Number=1,Type=Integer,Description="Total read depth">
##INFO=<ID=AF,Number=A,Type=Float,Description="Allele frequency">
##FILTER=<ID=LowQual,Description="Low quality">
##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
##FORMAT=<ID=AD,Number=R,Type=Integer,Description="Allelic depths">
##FORMAT=<ID=GQ,Number=1,Type=Integer,Description="Genotype quality">

## Data lines
#CHROM  POS      ID          REF  ALT  QUAL  FILTER  INFO                  FORMAT   SAMPLE
chr1    925952   rs2296290   G    A    1256  PASS    DP=87;AF=0.52;QD=14.4 GT:AD:GQ 0/1:42,45:99
chr17   43045629 rs28897696  G    A    .     .       .                     GT       0/1

Genotype Fields Explained

GT 0/0 - homozygous reference (both alleles = REF)
GT 0/1 - heterozygous (one REF, one ALT) - expected in germline Mendelian
GT 1/1 - homozygous alternate (both alleles = ALT)
GT 0/2 - heterozygous for second ALT allele at multi-allelic site
GT ./. - missing genotype (insufficient coverage or filtered)
GT 0|1 - phased: pipe separates maternal|paternal allele
GQ (Genotype Quality): Phred-scaled confidence; GQ<20 = unreliable genotype
AD (Allelic Depth): reads supporting REF,ALT (e.g., 42,45 for balanced het)
PL (Phred-scaled Likelihoods): 0/0,0/1,1/1 genotype likelihoods; lowest = called GT

VCF Manipulation with bcftools

code

# Count variants by type
bcftools stats sample.vcf.gz | grep "^SN"

# Filter PASS variants only
bcftools view -f PASS sample.vcf.gz -o pass.vcf.gz -O z

# Extract specific region
bcftools view sample.vcf.gz chr17:43000000-44000000

# Merge multiple VCFs
bcftools merge -m all *.vcf.gz -o merged.vcf.gz -O z

# Normalise indels (left-align and split multi-allelics)
bcftools norm -m-any -f hg38.fa sample.vcf.gz -o norm.vcf.gz

# Extract sample-level fields to TSV
bcftools query -f '[%CHROM\t%POS\t%REF\t%ALT\t%GT\t%GQ\n]' sample.vcf.gz

Variant Calling Variant Annotation