Free browser tool

Free raw DNA file checker

Direct answer: This checker can inspect an original raw-DNA ZIP or text export for a recognizable provider header, structural genotype rows, aggregate call rate, and a genome build only when the file declares one. It runs locally and reports no individual DNA calls.

Tool behavior and provider documentation reviewed 26 August 2026

Runs in this browser

Check your raw DNA file

Choose the original ZIP, TXT, CSV, or TSV export. The checker reports only file-level structure and aggregate counts. Maximum compressed or expanded size: 64 MB.

Choose a raw DNA file

or drop it here

Nothing is sent to DNA Layer by this checker.

Need to find an exact rsID?

The checker intentionally returns only aggregate structure. To reproduce a specific chromosome, position, and genotype row without uploading the file, use the local raw DNA viewer and rsID lookup. It does not attach gene, condition, or health meaning.

What it checks

Recognizable 23andMe, AncestryDNA, MyHeritage, TellMeGen, or generic table structure
Structural genotype-row count and aggregate no-call rate
GRCh38, GRCh37/hg19, or NCBI36/hg18 only when declared in the file header
Whether a readable TXT, CSV, or TSV table exists inside the original ZIP

What it does not check

No health, ancestry, trait, carrier, or medication interpretation
No validation of every genotype or inference of chip version from row count
No promise that a valid-looking file covers a particular DNA Layer finding
No VCF, CRAM, BAM, PDF, screenshot, or family-tree GEDCOM support

Exactly what stays on your device

The selected file is passed directly to a browser worker, checked there, and discarded when the worker finishes or the page closes. The checker does not send or retain the filename, ZIP entry name, file bytes, genotype rows, individual calls, aggregate counts, or result. It also does not infer sex or track checker interactions.

DNA Layer’s ordinary site analytics may record a page view. They do not receive the selected file or checker result. If you later choose to analyze your DNA, that is a separate, explicit flow described in the complete DNA data-flow guide.

How common raw DNA exports differ

Provider or file	Recognized columns	Accepted input
23andMe	rsid, chromosome, position, genotype	ZIP, TXT
AncestryDNA	rsid, chromosome, position, allele1, allele2	ZIP, TXT
MyHeritage	RSID, CHROMOSOME, POSITION, RESULT	ZIP, CSV, TXT
TellMeGen	rsid, chromosome, position, genotype	ZIP, TXT
Generic table	Four or five structural columns	Readable, but provider unverified
VCF	Variant Call Format	Recognized as unsupported

Provider interfaces and exports change. Keep the original download unchanged and use the current steps for 23andMe, AncestryDNA, or MyHeritage if you need a new copy.

This table and the checker share one versioned compatibility source. Reuse the JSON reference or download the CSV. It records implemented structures and limitations, not a claim that provider formats never change.

How to read each checker result

Format: A provider name appears only when the file header declares it. A generic four- or five-column table remains provider-unverified.
Structural rows: Rows with an rsID-like marker, chromosome, numeric position, and a structurally readable genotype. This is not a genotype-by-genotype accuracy test.
Aggregate call rate: The share of recognized rows with a call instead of a no-call such as -- or 0. It should not be compared across providers as a clinical-quality score.
Genome build: GRCh38/hg38, GRCh37/hg19, or NCBI36/hg18 only when declared in the header. The checker never guesses a build or chip version from row count.

Synthetic row structures

These placeholders contain no real or customer DNA. They show why AncestryDNA uses five fields while the other common checker paths use four.

23andMe-style

rsid  chromosome  position  genotype
rs000000  1  100000  AA

AncestryDNA-style

rsid  chromosome  position  allele1  allele2
rs000000  1  100000  A  A

MyHeritage-style CSV

"RSID","CHROMOSOME","POSITION","RESULT"
"rs000000","1","100000","AA"

Fix common raw DNA file problems

Not recognized: return to the provider and download the raw DNA or DNA Data export—not a PDF, screenshot, account archive, or family-tree GEDCOM.
Unusually small: use the complete original download. Email attachments, partial copies, and spreadsheet exports can omit rows.
Low aggregate call rate: re-download before assuming the file is damaged. If the original repeats the result, the checker cannot determine why.
Build not declared: do not guess from provider or marker count. A downstream analysis must handle coordinate identity explicitly.
Provider unverified: the table structure is readable, but the header does not prove where it came from. Keep the original provider copy.
VCF, CRAM, or BAM: preserve the sequencing file; do not rename or convert it merely to pass this checker. Those upload paths are not live in DNA Layer.

A peer-reviewed consumer-genotype quality-control study likewise separated basic format recognition from assembly, completeness, strand, and other sanity checks. This browser utility intentionally performs only the first, narrow layer.

A readable file is not a medical result

This utility checks structure, not clinical accuracy. Consumer files cover selected positions, and a low no-call rate does not mean the file contains every relevant position. Never use this result to diagnose, treat, or change medication. Consequential findings may need confirmation in an appropriate clinical laboratory.

Raw DNA file checker questions

Does this raw DNA file checker upload or store my file?

No. The checker reads the selected file inside a browser worker and returns aggregate structural results. It does not send the filename, file bytes, genotype rows, individual calls, aggregate counts, or checker result to DNA Layer or analytics. Ordinary site analytics may still record that this page was viewed.

Can I check the original 23andMe or AncestryDNA ZIP?

Yes. The checker can inspect an original ZIP directly, as long as the compressed file and candidate expanded DNA table are each within the 64 MB safety limit. It also accepts TXT, CSV, and TSV genotype tables.

What does aggregate call rate mean?

It is the percentage of structurally recognized genotype rows that contain a call rather than a no-call marker such as --. It is a basic file-level signal, not a measure of clinical validity or complete genome coverage.

Does a valid result mean my DNA file is clinically complete?

No. A recognizable table may still omit positions needed for a particular finding, use a different chip version, or contain calls that require confirmation. This checker does not interpret health, diagnose conditions, or verify every genotype.

Why does the checker reject VCF files?

DNA Layer's current consumer upload paths are designed for supported genotyping-array exports, not whole-genome or whole-exome VCF files. VCF support is not live, so the checker identifies that format without implying compatibility.

Format and interpretation sources

Recognize the file? See what its tested markers can support.

Review supported upload paths