¶ Guide to Interpreting and Explaining DNA Relationship Results
Source folder: Manuals
Source file: Guide to Interpreting and Explaining DNA Relationship Results.pdf
File type: PDF document
Back to folder index
Guide to Interpreting and Explaining DNA Relationship
Results
- Introduction / What This Report Shows
- Every relationship/paternity/kinship DNA report is based on a statistical
comparison of genetic markers (usually STR loci) between the parties tested.
- The output is not a simple "yes/no" in many cases; instead, you'll often see a
probability or likelihood ratio that quantifies how strongly the genetic
evidence supports a particular relationship hypothesis over the alternative
(i.e. "not related in that way").
- The report may also include intermediate metrics (e.g. Paternity Index (PI),
Combined Paternity Index (CPI), kinship index, etc.) plus a conclusion
statement (e.g. "not excluded," "excluded," or "relationship supported / not
supported / inconclusive").
- Key Terms & Concepts (with plain-language definitions)
Term Definition / Plain Explanation
Allele / Locus A specific position on a chromosome where
variation is measured (one "marker").
Paternity Index (PI) At one marker, how many times more likely the
observed match is if the tested person is the
biological parent vs. a random, unrelated person.
Combined Paternity Index The product of the PIs across all markers; used to
(CPI) derive the final probability.
Probability (or Probability The percentage reflecting how likely the tested
of Relationship, PRI / POP relationship is, based on the genetic evidence.
/ etc.)
Excluded The tested relationship hypothesis is ruled out by
the data (i.e. mismatch(s) at loci).
Not Excluded or Included The data are consistent with the tested
/ Supported relationship (though probabilistic).
Likelihood Ratio (LR) / How many times more likely it is that the tested
Kinship Index relationship is true vs. false (similar in concept to
CPI in relationship tests).
Inconclusive / Borderline The result is neither strongly supportive nor
strongly exclusionary; additional testing or
participants may be needed.
- Interpreting the Results: What the Numbers Mean
Paternity / Parent-Child Tests (most "definitive" cases)
- If a man is excluded, that means the data rules out biological paternity
(probability = 0 %).
- If a man is not excluded, you'll see a high probability (e.g. 99.9 % or higher
in many labs) to indicate near certainty. For instance, EasyDNA states that
when a man is included, they expect a probability of 99.9 % (or more) for
their tests.
- It is possible to have a mismatch at one locus yet still reach a high overall
probability, due to rare mutations. Many labs treat a single mismatch as
acceptable if the rest of the data is strongly supportive.
- Always check the conclusion statement (excluded / not excluded) in the
report - that is the lab's summary verdict.
Relationship Tests (Sibling, Avuncular, Grandparentage, etc.)
These are more complex, with less certainty, because you are inferring relationships
rather than testing a direct parent-child bond.
Example Interpretive Framework (from EasyDNA / general practice):
- High probability (e.g. ≥ 90 %): the tested relationship is supported by
DNA evidence (e.g. "the sibling relationship is supported")
- Intermediate / midrange (e.g. 9 % - 89 %): the result is inconclusive
- the evidence is not strong enough either way; adding more participants
(e.g. known parent, other relatives) may help
- Low probability (below ~ 9 %): the tested relationship is not supported
by DNA evidence
- For example, EasyDNA suggests that for sibling tests:
- ≥ 90 % → supported
- 9 % - 89 % → inconclusive
- < 9 % → not supported (as you noted)
However, these cutoffs are lab-specific and somewhat arbitrary, so always refer to
that lab's "interpretation" section.
For avuncular / aunt-uncle / grandparentage tests, the same principle applies: the
report will show a probability (or kinship index), and an interpretive statement from
the lab (e.g. "relationship supported / not supported / inconclusive").
Important nuance: even if a test shows less than the lab's "supported" threshold,
that does not always strictly rule out the relationship, especially in distant
relationships. E.g. the U.S. Citizenship & Immigration Services (USCIS) guidance
notes that valid full-sibling relationships may sometimes present with < 90 %
probability.
- What "Inconclusive" Means & When to Add More Testing
When results are inconclusive, it means the current data is insufficient to strongly
support or refute the relationship hypothesis. Possible reasons:
- The tested markers do not give enough statistical power
- The relationship is more distant (e.g. grandparent, uncle, etc.), which
naturally yields a weaker signal
- The samples are degraded or of low quality
- You've tested too few relatives
What you can do:
- Add a known parent (if available) - this often dramatically improves the
discriminating power
- Test additional relatives (e.g. siblings, nephews/nieces, aunts/uncles)
- Use more markers (if the lab offers an extended panel)
- Request the lab's expert review or additional research if they offer that
When explaining to clients, you might say:
"At present, the evidence doesn't lean decisively one way or the other. Adding a
parent or another close relative often helps tip the balance."
- Special Cases & Relationship Types
Here are some specifics you can include in your guide for common test types:
Sibling Testing
- Distinguish full-sibling vs half-sibling scenarios
- Emphasize that siblings don't inherit identical DNA, so variation is normal
- The probability / kinship index must be interpreted in context (e.g. using the
lab's thresholds)
- If mother's DNA is included, it helps filter maternal contribution, improving
clarity
Avuncular (Aunt/Uncle) Testing
- Probability expresses the chance that Person A is biologically an aunt/uncle to
Person B
- More tenuous than sibling or parent-child tests, so threshold for "support"
tends to be lower or more cautious
- Lab will state "relationship supported / not supported / inconclusive"
alongside the probability
Grandparentage Testing
- Similar principles: you'll get a probability that the tested grandparents are
biologically related
- Used when a parent is unavailable for testing
- Again, lab thresholds and interpretive statements accompany the number
mtDNA / Y-Chromosome / X-Chromosome Tests
- These are lineage-based tests rather than relationship probability tests
- They often result in a match / no match rather than a percentage, since they
trace direct maternal (mtDNA) or paternal (Y-chromosome) lines
- If two people share a Y-chromosome profile, it strongly suggests a common
paternal lineage (if in the same male line)
- Mitochondrial DNA (mtDNA) is passed unchanged along the maternal line, so
matching mtDNA indicates a shared maternal ancestor (not necessarily a
close relationship)
- Limitations, Caveats & Factors That Can Affect Accuracy
It's vital to communicate to clients that DNA relationship results are powerful, but
not infallible. Key caveats:
- Statistical & population assumptions: calculations rely on allele frequency
databases; different labs/populations may yield slightly different probabilities.
- Mutations: a genuine biological relationship may still show a mismatch at a
locus due to mutation.
- Related potential fathers / close relatives: if two possible fathers are
brothers, the test may struggle to distinguish them. Some labs require
additional testing in such cases.
- Sample quality / contamination / degradation: poor DNA samples can
reduce reliability
- Laboratory protocols / number of markers: more markers = greater
discriminating power
- Prior / background probabilities: the reported probability is conditional on
the assumptions built into the lab's statistical model
- "Inclusion" does not equal "100 %": no test claims absolute certainty -
even a very high percentage (e.g. 99.99 %) is still a probabilistic statement
- Interpretation thresholds differ by lab: always refer to the lab's own
"interpretation" legend
- Example Interpretations (Plain Language)
These examples show how to read typical result statements and translate them into
clear, client-friendly explanations.
Always begin by confirming what type of analysis* was done (paternity, sibling,
grandparentage, etc.) - each has slightly different interpretation thresholds.
Test Type Report Statement How to Explain to Client
/ Example Data (Plain Language)
Paternity "Alleged father is not "The DNA analysis shows an
excluded as the extremely high probability that the
biological father. tested man is the biological father.
Probability of paternity: In other words, the genetic
99.999 %." evidence fully supports paternity."
Paternity "Alleged father is "The DNA profiles do not match on
excluded as the several key markers, so the tested
biological father due to man cannot be the biological
mismatches at three father."
loci."
Sibling (Full) "Probability of "The analysis supports that the two
relationship: 93 % individuals are full siblings, sharing
(Full-sibling both biological parents."
hypothesis)."
Sibling (Half) "Probability of half- "This result leans toward a half-
sibling relationship: 84 sibling relationship, but it is not
%." conclusive. Adding one parent's
sample would help confirm."
Sibling "Probability: 47 %." "The evidence is balanced and does
(Inconclusive) not clearly confirm or exclude a
sibling relationship. Further testing
is recommended."
Grandparentage "Probability of "The result supports that the
relationship: 95 %." tested person is the biological
grandparent of the child."
Grandparentage "Probability: 25 %." "This probability is too low to
support a biological grandparent
relationship. It may indicate no
biological connection or that the
relationship is more distant."
Avuncular "Probability of "The DNA evidence moderately
(Aunt/Uncle) relationship: 82 %." supports that the tested person is
a biological aunt or uncle, though
additional relatives could improve
certainty."
Y-Chromosome "Y-STR profiles match "The two male individuals share
at all tested loci." the same paternal lineage - they
likely descend from the same male
ancestor."
Y-Chromosome "Profiles differ at "The males do not share the same
multiple loci." paternal line."
mtDNA "mtDNA sequences are "The individuals share the same
(Mitochondrial) identical." maternal lineage - they likely
descend from the same maternal
ancestor."
mtDNA "mtDNA profiles differ." "There is no shared maternal
(Mitochondrial) lineage between the tested
individuals."
- "Why isn't it just yes or no?"
- "Can a mutation cause a mismatch?"
- "What if the result is inconclusive?"
- "Will adding my parent/aunt/uncle help?"
- "Can two brothers both be possible fathers in a paternity test?"
- "How reliable is this test?"
- "What happens to my DNA sample after testing?"
- "Can this be used in court / for immigration?" (discuss chain of custody,
accreditation)
- Suggestions for How You (as Advisor) Should Explain to Clients
- Use plain language - avoid technical terms unless you immediately define
them
- Start with the bottom line (conclusion) first, then walk through how the
probabilities support it
- Use analogies (e.g. "the test says that this relationship is X times more likely
than not," or "if this were a courtroom, the evidence strongly supports/does
not support the claim")
- Emphasise uncertainty: "This is a probability, not absolute proof"
- When results are inconclusive, frame it as "this is an invitation to add more
data, not a failure"
- Be empathetic: clients may have emotional stakes in these results
- Offer to walk them step by step through the report, pointing to each section
(markers, indices, conclusion)
- Encourage clients to ask questions if anything is unclear