All studies
DNA·October 29, 2026·4 min read

How to Convert a VCF to PLINK Format Without the Command Line

Convert a whole-genome VCF or .vcf.gz into PLINK .ped and .map files free in your browser, with no bcftools or PLINK install.

You can convert a whole-genome .vcf or .vcf.gz into PLINK .ped and .map files for free in your browser at the free VCF-to-PLINK converter. You do not need bcftools, PLINK, or a terminal. The file is read on your device and never uploaded.

What PLINK .ped and .map actually are

PLINK 1.x text format is two files that travel together. The .map file describes each variant: chromosome, marker ID, genetic distance and base position. The .ped file has one row per person. The first six columns are family ID, individual ID, paternal ID, maternal ID, sex and phenotype. After that, each variant uses two columns for the two alleles. For example, a SNP at chromosome 1, position 100 with alleles A and G appears in the .map as 1 rs123 0 100, and in the .ped as two columns A G. The converter writes PLINK 1.x text, zipped.

How do I convert a VCF to PLINK format online free?

Go to the converter page, drop your .vcf or .vcf.gz, and choose PLINK as the output format. The page reads plain gzip and BGZF VCFs, including the files bcftools and GATK write. It writes a zipped .ped and .map pair on your device. There is no account, no upload and no command line.

VCF to PLINK without the command line
Whole-genome VCF or .vcf.gz
GT fields read locally, optional GRCh37 lift
PLINK 1.x text
.ped + .map, zipped download
3 to 5M
variants in a typical whole-genome VCF
2
files downloaded: .ped and .map
0
uploaded to any server

How the conversion works, row by row

Each VCF row has chromosome, position, ID, REF, ALT, FILTER and the sample column. The converter reads the GT field in that sample column. If GT is 0/1, it writes the REF and ALT letters. If GT is 1/1, it writes ALT twice. A ./. or .|. no-call stays missing in the .ped. PLINK text has no phase, so phased genotypes are written the same as unphased. The result is a plain genotype copy, not a new call.

What carries over

The converter copies each genotype from the GT field exactly. A no-call stays a no-call. It does not invent genotypes at positions your VCF does not list. An all-sites VCF or gVCF that explicitly calls reference positions will carry those; a standard whole-genome VCF usually lists only where you differ from the reference. The raw converter statement says PLINK 1.x text, genotypes copied, no-calls stay no-calls, and an optional GRCh38-to-GRCh37 lift. It does not claim filtering by variant type, FILTER status, or contig. Check the live page for the current row-handling details before you rely on any filtering.

What carries from a VCF into PLINK .ped/.map
VCF contentPLINK output
Variant with an rsID already in the VCF ID columnID column copied to the .map marker ID; two .ped allele columns
Variant with no rsID in the VCF ID columnNo rsID is assigned; the .map marker ID may stay missing
GT 0/1Writes REF and ALT
GT 1/1Writes ALT twice
No-call ./ or .|.Stays missing in .ped
Phased GTWritten as unphased in PLINK text
GRCh38 VCF with lift selectedPositions mapped to GRCh37; unmapped positions omitted

The converter copies GT calls as PLINK alleles. No-calls stay missing. Optional GRCh38-to-GRCh37 lift may omit unmapped positions.

What about GRCh38 and GRCh37?

By default, positions are copied exactly as they are. If your VCF is on GRCh38 and the downstream PLINK tool expects GRCh37, the converter has a browser-based lift. It uses Ensembl's GRCh38-to-GRCh37 chain, flips alleles where the chain crosses strands, and leaves out positions it cannot place. It only moves positions when you select that option. GRCh37 and GRCh38 are not the same coordinate system, so this is not a simple offset.

Does the site upload my genome?

No. The browser opens the file on your device and parses it locally. The page may download a public chain file or SNP list it needs for the conversion, but nothing from your VCF is sent back. The service counts that a conversion ran and the formats and outcome, never the file, its name or any genotype.

What people get wrong

  • PLINK is one file. It is not. You need both the .ped and the .map for any tool that reads PLINK text.
  • The output is binary PLINK. The converter writes PLINK 1.x text. Some tools ask for binary .bed, .bim and .fam files. Those still need desktop PLINK to create from the text pair.
  • Everything in the VCF carries over. The converter copies GT calls and no-calls, but the raw statement does not promise filtering by variant type, FILTER status, or contig. Check the live page for current details.
  • No rsIDs is fatal. A raw sequencing VCF sometimes has . in the ID column. The VCF-to-PLINK converter page does not promise to name unnamed VCF rows from a known SNP list. A VCF without rsIDs may need bcftools annotate against dbSNP before conversion, or check the live tool page for current behavior.

After you have the PLINK pair

The full free DNA report can read the same VCF directly for ancestry across 15 world regions, traits, medication genes and both family lines. That is often the fastest way to see whether your VCF is worth converting for more specific tools.

New to raw DNA files? Start with the full guide: How to Download & Analyze Your 23andMe Raw Data Free

Written by

HealthOS Research

Keep reading