← All guides

Analysis tools

Once you have a match list, these tools help you make sense of it: group matches into family lines, read your deep paternal and maternal ancestry, and check that a kit is healthy.

🧩 Clustering

IntermediateOpen tool

Groups your matches into clusters of people who share DNA with each other, not just with you. Each cluster usually represents one ancestral line (a grandparent branch, for example).

When to use it: After One-to-Many, when you have more matches than you can track and need to split them into paternal and maternal sides.

How to use it
  1. Choose a kit and a minimum shared cM (20 cM is the classic Leeds-method threshold).
  2. Run it; each cluster lists its members.
  3. Identify one known relative in a cluster — everyone else in that cluster is likely on the same branch.

Reading the result:

  • A tidy result is 4–8 clusters, roughly matching your grandparents.
  • Clusters that share members can indicate related lines or endogamy.
Keep in mind:
  • Endogamous populations (including many Roma families) can merge clusters, because distant relatives share more DNA — and often double (IBD2) segments — than usual. One runaway cluster may swallow a large share of the matches.
  • If that happens, raise the minimum cM (e.g. to 50) so only the strongest matches define the clusters, and use the clustering page's max-cluster-size option to break up a mega-cluster.
  • Only the strongest matches are used; very distant matches are ignored by design.

🌳 AutoKinship

AdvancedOpen tool

Builds and scores possible family trees from your matches alone — no existing tree required. It tries every way the matches could sit around you (as parents, siblings, cousins, and so on) and ranks the arrangements by how well the observed cM totals fit.

When to use it: When you have several matches whose relationship to each other is unclear, or you are working without a documented tree (adoptees, unknown parentage).

How to use it
  1. Choose your kit.
  2. Let it enumerate the trees; it shows the top-scoring arrangements with an SVG diagram.
  3. Lock any relationships you already know so the tree respects them.

Reading the result:

  • The highest-scoring tree is the most consistent with the DNA.
  • Same-person duplicate kits are merged automatically and noted.
Keep in mind:
  • A predicted tree is a hypothesis. Confirm placements with WATO, records and additional testers.
  • Endogamy inflates apparent cM and can confuse generation inference — in Roma families expect the predicted relationships to skew too close.

🧬 Y-DNA

IntermediateOpen tool

Reads the Y-chromosome SNPs in a male kit and walks them down a haplogroup tree to name your deep paternal line — the same line your surname usually follows.

When to use it: To test whether two males share a paternal ancestor, or to place your paternal line in a haplogroup.

How to use it
  1. Choose a male kit (female kits have no Y chromosome and are flagged as no-data).
  2. Read the haplogroup and the branch path that produced it.

Reading the result:

  • A letter-and-number code such as R-U106 or J2a2b1 — the deeper the branch, the more specific.
  • A "Roma founder line?" flag appears when the kit falls on a haplogroup associated with Roma ancestry (for example the H branch).
Keep in mind:
  • Chips only test a few thousand Y positions, so the haplogroup is often shallower than a dedicated Y-DNA test would give.
  • Two males who share a haplogroup may still be separated by many generations.

Haplogroup tree data: 23andMe yhaplo / ISOGG 2016.

🪡 mtDNA

IntermediateOpen tool

Reads the mitochondrial-DNA positions present in your chip and walks PhyloTree Build 17 to name your maternal-line haplogroup.

When to use it: To place your direct maternal line, or to check whether two people share a maternal ancestor.

Reading the result:

  • A haplogroup such as H or H3af, plus the markers that resolved it.
  • Flags for Roma-associated branches (M5a1, M18, M35b) are shown when present.
Keep in mind:
  • Chip coverage limits depth. A chip tests only a few hundred of the 16,569 mtDNA positions, and key markers are often missing, so two people with the same maternal line can resolve to different depths (e.g. one to H, one to H3af) purely from which markers their file happened to include.
  • The page labels weak or coverage-limited results so you know when to be cautious.

PhyloTree Build 17 (van Oven, 2015).

❌ X-DNA

IntermediateOpen tool

Shows how much X-chromosome DNA your kits share, with an inheritance explainer and a top-matches list.

When to use it: When a match does not show up on the autosomes as expected, or to follow a specific maternal or paternal X path.

Reading the result:

  • A mother and son share a near-complete X; two full sisters share about half.
  • A male has one X, inherited from his mother, so two males in a row break X transmission.
Keep in mind:
  • The X has no haplogroups and recombines in females, so it complements — but does not replace — autosomal matching.

🌍 Admixture / Ethnicity

IntermediateOpen tool

Estimates your ancestry composition from reference populations, with tabs for admixture calculators, a Romani route and community match, and G25 similarity.

When to use it: To see the South Asian, Middle Eastern and European mix that often signals Roma ancestry, and to compare against Roma reference communities.

How to use it
  1. Choose a kit (or paste Global25 coordinates where allowed).
  2. Run the calculator; the first run takes about a minute and is cached afterwards.
  3. Open the Romani tab for the route breakdown and the closest Roma community references.

Reading the result:

  • Component bars show the modelled proportions (e.g. North Sea, North Atlantic, Italian, Iberian).
  • The Romani tab contrasts a South-Asian component against host-population components — a distinctive Roma signal looks like South Asian plus Middle Eastern plus European together.
Keep in mind:
  • Estimates depend on the reference panel and model; treat percentages as approximate.
  • Attribution: Eurogenes calculators (Davidski).

🩺 Kit Diagnostics

BeginnerOpen tool

Reports the health of an uploaded kit: number of SNPs, coverage, no-call rate, heterozygosity, detected sex and chromosome coverage.

When to use it: When matches look wrong, when a kit has very few SNPs, or before trusting a surprising result.

Reading the result:

  • A healthy modern chip has hundreds of thousands of SNPs with a low no-call rate.
  • Very low coverage or a strange heterozygosity value can explain poor matching.

Get more out of your matches

The family-tree guide shows how to attach confirmed matches and use WATO to place unknown ones.

Upload your raw DNA