Navigating an NPE Discovery
Navigating an NPE Discovery: The Complete Genetic Genealogy Manual
A step-by-step manual for resolving Non-Parent Expected discoveries, identifying unknown biological parents and mastering autosomal DNA evidence.
When the Tree and the DNA Disagree
Conceptual illustration. The documented line (left) does not match the DNA; matches who share DNA with each other (right) point toward the unknown biological family.
1. Understanding the NPE Discovery
Taking a DNA test with AncestryDNA, 23andMe, MyHeritage or FamilyTreeDNA often starts as curiosity about ethnicity or a wish to confirm the paper trail. For tens of thousands of testers every year, the match list instead reveals a close biological match who fits nowhere on the documented tree, or the absence of expected paternal or maternal lines.
In genetic genealogy this is called an NPE: Non-Parent Expected (historically Non-Paternity Event). It occurs when a person's biological parentage differs from what they believed or were raised to understand. It covers unexpected paternity, undisclosed adoption, donor conception, secret step-parentage, and misattributed parentage further back, such as an unknown grandfather.
Resolving an NPE takes two things at once: scientific objectivity and careful data management, and emotional resilience and ethical boundaries. This manual follows the protocol used by professional genetic genealogists and volunteer Search Angels.
📸 Crucial First Action: The Evidence Preservation Protocol
Do not message close matches before securing your evidence. Panic can lead people to make profiles private or delete accounts and trees. Once data is gone, clues are lost for good. Do these four things first:
2. Privacy Infrastructure & Psychological Preparation
Set up a safe research environment before the technical work begins. It protects you and relatives who may not know the story.
Dedicated research email
Create a fresh address (ProtonMail, Gmail) without your name, birth year or recognizable handles.
Private research trees
Keep working "Quick & Dirty" trees private and unsearchable so living candidates are not notified before you are ready.
Profile pseudonyms
On GEDmatch or MyHeritage, consider initials or a neutral research alias while you test hypotheses.
Emotional pacing
NPE discoveries can cause identity disruption. Pace the work, keep a research journal of facts, and lean on support networks.
3. Rule Out Misinterpretations & Decode Shared cM Ranges
Before concluding an NPE exists, rule out technical errors, administrative mix-ups and misread centimorgan (cM) values. Recombination is random, so shared cM values overlap between relationship categories. A common early error is mistaking a half-sibling for an aunt, uncle, grandparent, niece or nephew: all fall in the same range (about 1,300 to 2,300 cM).
How Much DNA Do Relatives Share?
The amber bar marks the ambiguous second-degree group. Age, X-DNA and shared matches separate half-sibling from aunt, uncle or grandparent.
| Relationship | Average cM | Typical range | FIR / HIR | Key marker |
|---|---|---|---|---|
| Parent / child | ~3,480 | 3,300 – 3,700 | 100% HIR, no FIRs | One allele shared at every locus across the genome. |
| Full sibling | ~2,550 | 2,200 – 3,400 | ~25% FIR, ~50% HIR | FIRs on multiple chromosomes. |
| Half-sibling | ~1,750 | 1,300 – 2,300 | HIR only, zero FIR | Separate from aunt/uncle by age, X-DNA, shared matches. |
| Aunt / uncle / grandparent / niece / nephew | ~1,750 | 1,300 – 2,300 | HIR only, zero FIR | Grandparent shares ~25% across all chromosomes. |
| 3/4 sibling | ~2,150 | 1,800 – 2,600 | Small scattered FIRs | One shared parent; the other parents are full siblings. |
| 1st cousin / half-aunt / great-aunt | ~850 | 500 – 1,200 | ~12.5% shared | Shared great-grandparent or half-grandparent couple. |
| 1C1R / half 1st cousin | ~425 | 200 – 650 | ~6.25% shared | Shared 2nd great-grandparent couple. |
| 2nd cousin / half 1C1R | ~225 | 75 – 360 | ~3.1% shared | Shared 2nd great-grandparents. Key for clustering. |
Pro tip: test exact cM values and segment counts with the Shared cM Project or the Genetic Voyage DNA Relationship Probability Calculator.
FIRs vs. HIRs: Full Siblings or Half-Siblings?
Full siblings share both parental copies in some regions. Half-siblings never do, unless their shared parent's family is related.
🔬 Technical Deep Dive: FIRs, HIRs and X-DNA
- Half Identical Regions (HIRs): one chromosome copy matches. All relatives except parent-child and full siblings share only HIRs.
- Fully Identical Regions (FIRs): both copies match. Full siblings share ~25% FIRs, half-siblings 0%. Zero FIRs between two ~1,700 cM matches means half-siblings or aunt/uncle, not full siblings.
- X-chromosome rules: males get their only X from their mother; females get one from each parent. A father passes his X intact to every daughter, so paternal half-sisters share a full, unbroken X (~182 cM), while maternal half-sisters may share only recombined segments. Two males can never share paternal X-DNA.
4. The Cross-Database Passport Protocol
Never rely on one database. A 1st cousin who could solve your case may have tested on MyHeritage while you tested on Ancestry. Raw data upload extends your reach.
| Platform | Size | Upload? | Key feature for NPE research |
|---|---|---|---|
| AncestryDNA | 25M+ | ❌ Test here | Largest database, tree network, SideView™ parental phasing. |
| 23andMe | 14M+ | ❌ Test here | Explicit FIR/HIR breakdown, chromosome browser, haplogroups. |
| MyHeritage | 7M+ | ✅ Free upload | Strong European database, AutoClusters, segment browser. |
| FamilyTreeDNA | 1.5M+ | ✅ Free upload | Chromosome browser, Y-DNA (Big Y-700) and mtDNA. |
| GEDmatch | 1.5M+ | ✅ Free upload | Cross-company comparison, one-to-one segments, AYPR tool. |
See How to Download Raw DNA Data & Upload to Free Databases.
🧬 Mandatory Diagnostic: GEDmatch “Are Your Parents Related?” (AYPR)
Upload your raw DNA to GEDmatch and run AYPR. It scans for Runs of Homozygosity (ROH), regions where your maternal and paternal chromosomes match each other.
- No significant ROH: your parents are likely unrelated.
- Large continuous ROH blocks (for example 100+ cM in total): your parents share recent ancestry, from 2nd/3rd cousins to closer.
- Why it matters: endogamy or consanguinity distorts relationship estimates on the mystery line.
5. Isolating the Mystery Line: Step-by-Step Clustering
Once the NPE is confirmed and matches are consolidated, separate your match list into ancestral clusters and isolate the unknown biological line.
From Match List to Mystery Line
Four grandparent-level clusters (amber) emerge from the filtered mystery set.
Phase 1: Separate known-parent matches
Color-label every match who shares DNA with your known parent's relatives and exclude them.
Phase 2: Isolate mystery matches
Filter for matches of 40 – 400 cM who share zero DNA with the known parent's line.
Phase 3: The Leeds Method
Sort mystery matches (90 – 400 cM) into four color columns, one per biological grandparent line.
Phase 4: Run AutoCluster
Use AutoCluster or MyHeritage in-common-with matrices for visual ancestral branches.
🌳 Building "Quick & Dirty" Mirror Trees
Most matches have tiny trees or none. Build research trees for them yourself:
- Select a mystery cluster: 3–5 matches who share DNA with one another (for example 210, 140 and 90 cM).
- Trace backward: use census records, birth/marriage/death indexes, obituaries and public trees.
- Find the MRCA: the couple where the trees merge. You descend from them too.
- Trace forward: follow all their descendants to the generation of your unknown parent.
- Audit candidates: list the people who were alive, of age and in the right place at the time of conception.
6. Testing Candidates with WATO & BanyanDNA
Once a candidate tree exists, test your position in it mathematically before drawing conclusions.
WATO (What Are the Odds?) v2
On DNA Painter, WATO lets you import the mystery tree, enter each match's cM, and place yourself as a hypothesis at every candidate position. It scores all matches together and gives a Score Ratio (for example, Hypothesis 1 is 4,500 times more likely than Hypothesis 2).
WATO vs. BanyanDNA Guide →BanyanDNA & Triangulation
For half-relationships, pedigree collapse, endogamy, or candidates who are brothers or cousins, BanyanDNA models probabilities across multi-person networks. Pair it with segment triangulation to confirm shared blocks.
Triangulation Guide →7. Targeted DNA Testing & Confirmation
When research narrows to one candidate branch (for example three brothers), targeted testing can settle it with minimal friction.
Suspected half-siblings
A child of the candidate parent should share ~1,750 cM. FIR/HIR status, or X-DNA between females, confirms it.
1st or 2nd cousins
If candidates are deceased, testing descendants of candidate uncles or aunts narrows WATO to one line.
Y-DNA (male seekers)
Y-37, Y-111 or Big Y-700 at FamilyTreeDNA follows the paternal line, and matches often reveal the biological father's surname.
mtDNA (maternal line)
Mitochondrial DNA passes from mothers to all children and confirms the direct maternal lineage.
8. Outreach Etiquette & Communication Templates
The person you contact may not know the family secret, or may feel protective of their privacy. Core rules:
- Never lead with dramatic assertions. "You are my half-sibling" in a first message triggers defensiveness and deleted accounts.
- Focus on shared genealogy. State your shared cM, mention a common ancestral connection, and ask to compare surnames or locations.
- Be concise. Keep the first message under 150 words.
- Keep boundaries. Never post living names or unverified claims on social media or open forums.
✉️ Outreach Script Library: 3 Message Templates
Scenario A: distant mystery match (2nd/3rd cousin, 90–250 cM)
Scenario B: close unknown match (half-sibling, aunt/uncle, 1st cousin, 400+ cM)
Scenario C: follow-up to a non-responsive match
9. Support Networks & Search Angel Resources
You do not have to carry this alone. Non-profit groups and volunteers offer free technical help and peer support:
NPE Friends Fellowship
Peer support groups, educational events and emotional resources for people who have had an NPE discovery.
DNA Angels (dnaangels.org)
Volunteer "Search Angels" who help adoptees, NPEs and donor-conceived people identify biological parents free of charge.
Severed Roots & DNA NPE Gateway
Peer-led online communities offering confidential space to process unexpected findings.
Donor Conceived Community
Support and education for people discovering donor conception.
Watershed DNA
Educational and coaching resources on genetic identity and unexpected DNA discoveries.
10. The Complete NPE Master Checklist
📋 Recap: Your Step-by-Step Action Plan
- Preserve evidence immediately. Download raw DNA, screenshot top matches, shared-match lists and trees before contacting anyone.
- Establish research privacy. Private email; private, unsearchable working trees.
- Rule out ambiguity. Verify cM totals, segment counts, FIR/HIR status and X-DNA patterns.
- Upload across databases. MyHeritage, FTDNA and GEDmatch for maximum match coverage.
- Run GEDmatch AYPR. Check for endogamy or consanguinity.
- Isolate the mystery line. Remove known-parent matches; apply the Leeds Method or AutoCluster.
- Build Quick & Dirty mirror trees. Trace clusters back to MRCAs, then all descendants forward.
- Test hypotheses in WATO / BanyanDNA. Calculate probabilities for each candidate position.
- Test targeted relatives. Candidate relatives, Y-DNA or mtDNA to confirm the line.
- Reach out with empathy. Neutral, non-confrontational templates only.
Continue Your Genetic Genealogy Research
Master the tools, methods and calculators that turn complex DNA match lists into documented family history.