mtDNA haplogroup finder

Mitochondrial DNA is inherited from your mother and passes down the maternal line almost unchanged, so like the Y chromosome it forms a single-line tree — but everyone has it, not only men. This tool reads the mitochondrial variants out of a 23andMe, AncestryDNA or MyHeritage export, or out of a whole-genome VCF from a sequencing kit, matches them against the published tree, and reports the branch your variants place you on, the specific variants that placed you there, and which branches below it your file could and could not test.

Your maternal haplogroup describes one thread: your mother's mother's mother, and so on. It is not an ethnicity and not a measure of your overall ancestry, which is spread across the rest of your genome. The mitochondrial genome is also small — a few thousand times smaller than the nuclear genome — so it resolves a lineage, not a population history.

Coverage sets the ceiling here too. Consumer chips test a limited selection of mitochondrial positions, so the branch you can be assigned is bounded by which of them your file contains; a broad assignment usually reflects a sparse file rather than a shallow lineage. Showing which downstream branches remain untestable is the honest form of that limit, and where the evidence is too thin the result says so rather than reaching for a deeper branch. A whole-genome VCF carries the full mitochondrial sequence, so it can reach a terminal branch a chip cannot. Runs need a free, email-verified account, and the uploaded file is deleted after the analysis.

Suppose a file comes back as H1, with the placing variants listed, several H1 subclades marked untestable, and no deeper call offered. The first calibration: broad maternal results are NORMAL from chips. The mitochondrial genome is tiny and consumer arrays test far fewer mitochondrial positions than Y positions, so most kits resolve to a major branch — H, H1, U5, K, T2 — and that is a correct answer at the data's resolution, not a failed test. The placing variants are the receipt: each is a defined mutation against the rCRS reference that your file carries in the derived state.

The untestable list tells you what full sequencing would add. H1 alone spans millions of living people; its named subclades (H1a, H1c and dozens more) subdivide it into lineages young enough to be genealogically interesting — and nearly all of them are defined by positions chips do not read. If your maternal line is the research target, that list is effectively the shopping guide for a full-sequence test.

Reading the branch itself follows the haplotree pages' discipline: a maternal haplogroup is your mother's mother's line and nothing else — not an ethnicity, not a percentage, and not evidence about the other thousands of ancestors at the same depth. The interesting genealogy is usually in the branch's geography, which the haplogroup atlas maps from ancient carriers and modern testers alike.

The community's classic free tool is James Lick's mthap, which reads chip extracts against the mitochondrial tree and reports best-fit haplogroups with marker lists — a genuinely good instrument that generations of hobbyists learned on. This tool covers the same ground with a maintained tree, explicit untestable-branch reporting, the same no-guessing rule where evidence thins, and server-side runs that delete the upload afterwards. For chip-grade input, expect the two to agree; disagreement usually traces to tree versions rather than either tool erring.

The real alternative is a different test, not a different reader: FamilyTreeDNA's full mitochondrial sequence reads all ~16,569 positions, resolves to the deepest named branch, and enables matching on exact sequences. It is the only way past the chip ceiling — no reader, ours included, can call subclades from positions a file never tested. The proportionate path: read your existing file free, see which branches stay untestable, and let that decide whether sequencing buys anything your question needs. Either way the maternal line is one thread; the autosomal tools carry the rest of the story.