Free VCF converter
Convert your VCF to 23andMe format
Some raw DNA tools only take a 23andMe-style text file. Drop a whole-genome .vcf or .vcf.gz and download your genotypes as rsid, chromosome, position and genotype, the layout 23andMe uses. Your VCF needs rsIDs in its ID column.
The conversion runs entirely in your browser. Your file is never uploaded, there is no account, and it is free. We count that a conversion ran and its outcome, such as “GRCh37, converted”, and never the file, its name or any genotype.
Choose or drop your VCF
.vcf · .vcf.gz · .zip
You may not need to convert at all
The free DNA report at /dna reads a whole-genome VCF directly: medication genes, inherited risk factors, traits, ancestry and both family lines. Your DNA story opens straight away with no account; the full report opens after a code sent to your email.
What the converter does
- It writes one row for each SNP your VCF lists with an rsID: the rsID, the chromosome (1–22, X, Y or MT), the position, and your two letters, or one letter where the VCF has a single call, as on the Y chromosome in men.
- It leaves out insertions, deletions and structural variants (a 23andMe-style file holds single letters only), rows the variant caller flagged as unreliable, and anything off the main chromosomes.
- It copies positions as they are and says which genome build they are on. It does not add rsIDs or convert between builds.
- It reads .vcf and .vcf.gz, including the BGZF files bcftools and GATK write, at any size the browser can stream.
Questions
How do I convert my whole genome VCF to 23andMe format?
Drop the .vcf or .vcf.gz on healthosx.com/dna/vcf-to-23andme. It writes every SNP your VCF lists with an rsID as rsid, chromosome, position and genotype, the 23andMe raw data layout, and gives it back as a .txt or .zip. It runs in your browser: free, no account, and the file is never uploaded.
My VCF has no rsIDs. Can I still convert it?
Not here: a 23andMe-style file is keyed by rsID. Annotate the VCF against dbSNP first, with bcftools annotate -c ID and a dbSNP file on the same genome build, then convert the result. The free report at healthosx.com/dna reads an unannotated whole genome for ancestry without converting it.
Why does the converted file have fewer rows than my VCF?
It keeps only single-letter changes (SNPs) that have an rsID. Insertions, deletions, structural variants, rows the variant caller flagged and contigs off the main chromosomes are left out, and the page counts each.
Does the converted file list the places where I match the reference?
Only where your VCF does. A whole-genome VCF usually lists where you differ from the reference, so the converted file does too. A gVCF or an all-sites VCF also lists matching positions, and those are written with the reference letters.
Which genome build is the output on?
The same as your VCF: positions are copied, not converted. 23andMe files are on GRCh37, so a tool that matches on position needs a GRCh37 VCF; tools that look markers up by rsID work either way.
Is my genome uploaded?
No. The file is read and converted in your browser, and the download is made on your device.
Can I convert a BAM or CRAM file to 23andMe format?
Not directly. BAM and CRAM files hold your aligned sequencing reads, not a list of genotypes. Variant calling (bcftools or GATK) turns them into a VCF first, and most labs can export a VCF if you ask. Then convert the VCF here.
Where can I use the converted file?
With tools built for genotyping chip files, which expect 23andMe’s layout. Before uploading it to GEDmatch, Read Your DNA or any other site, check that site’s current size and format rules: a converted whole genome holds millions of rows, far more than a chip file, and a GRCh38 VCF gives GRCh38 positions. You don’t need to convert for healthosx.com/dna, which reads the VCF directly.