A sibling DNA test report boils down to a handful of numbers that express how likely it is two people share one or both biological parents. The centerpiece is usually a likelihood ratio or a combined sibling index, a single figure that compares the odds of two competing explanations for the genetic data. Understanding what that number means, where its limits are, and what the rest of the report adds is straightforward once you know what you’re looking at, but most testing companies do a poor job explaining it, leaving people to stare at a page of statistics with real emotional stakes and no clear guide.
What the Likelihood Ratio Actually Means
The likelihood ratio (often abbreviated LR on your report) is the single most important number. It compares two hypotheses, usually stated something like “the tested individuals are biological siblings” versus “the tested individuals are unrelated.” The ratio tells you how many times more likely the DNA evidence is under one scenario compared to the other. An LR of 500, for example, means the genetic profiles you both carry are 500 times more likely to exist if you are siblings than if you are strangers.
The higher the LR, the stronger the evidence for a biological relationship. A value of 1 is perfectly neutral, meaning the DNA is equally consistent with both explanations. Values below 1 favor the “unrelated” hypothesis. Most laboratories consider an LR above 100 to be strong support for a sibling relationship, and values above 10,000 are considered very strong. A study using 21 genetic markers found that about 80% of known sibling pairs produced LR values of 10,000 or higher, while only about 48% reached that threshold when fewer markers (15) were used.1Legal Medicine. Sibling assessment based on likelihood ratio and total number of shared alleles using 21 short tandem repeat loci included in the GlobalFiler™ kit So if your report uses a modern panel with more markers, the numbers tend to be more decisive.
Some reports present a “sibling index” instead of, or in addition to, the likelihood ratio. These terms are often used interchangeably. The sibling index at a single genetic location tells you how much that one spot in the genome supports siblingship. Your report’s “combined sibling index” multiplies those individual values across every location tested, producing a single overall figure. That combined number is essentially the same thing as the LR.
The Probability Percentage
Most consumer-facing reports convert the likelihood ratio into a percentage, sometimes labeled “probability of relatedness” or “probability of siblingship.” This is the number people tend to fixate on, and for good reason: seeing “99.7% probability” feels much more intuitive than seeing “LR = 340.” But the percentage is not a simple translation. It depends on a prior probability, which is an assumption the lab makes before looking at the DNA about how likely the relationship was in the first place.
Most labs use a 50/50 prior by default, meaning they assume equal odds that the two people are siblings or unrelated before any DNA is examined. With that prior, an LR of 100 converts to roughly a 99% probability. An LR of 10 converts to about 91%. An LR of 1 gives you exactly 50%. If the lab used a different prior, the percentage shifts. This matters because the prior is not a fact derived from DNA; it is a modeling choice. If you have strong non-genetic reasons to believe a relationship exists (for instance, you share the same mother and a likely father), a 50/50 prior is actually conservative. If the possibility of siblingship came out of nowhere, a 50/50 prior might be generous. Some reports let you see what prior was used, often in a footnote or methods section. If yours does not, it almost certainly defaulted to 50/50.
The practical takeaway: trust the direction and the magnitude of the likelihood ratio more than the exact percentage. A probability above 99% with a high LR is strong evidence. A probability hovering around 50 to 70% means the test could not clearly distinguish between the two hypotheses, and you may need additional testing.
Full Siblings, Half Siblings, or Unrelated
Many sibling tests don’t just ask “related or not.” They compare three (or more) hypotheses: full siblings (same mother and father), half siblings (one shared parent), and unrelated. Your report may present separate likelihood ratios for each comparison, or it may show a ratio that directly pits full siblingship against half siblingship.
This distinction is harder to make than most people expect. Full siblings share, on average, about half their DNA. Half siblings share about a quarter. But these are averages, and real pairs scatter around them. Some half siblings happen to share an unusually large chunk, while some full siblings share less than average. One early study using just three genetic markers found that the sibling index correctly favored full siblingship in only 18 of 25 known full-sibling pairs, and the ratio was strong (above 100) in just 8 of those 25.2Transfusion. Distinguishing full siblings from half-siblings in limited pedigrees Modern panels with 20 or more markers perform much better, but the fundamental challenge remains: telling full siblings from half siblings is genuinely more difficult than confirming a parent-child relationship.
If your report gives a combined index or LR for “full sibling vs. half sibling” rather than “sibling vs. unrelated,” pay attention to whether it exceeds common thresholds. An LR strongly above 1 favors full siblingship; an LR well below 1 favors half siblingship. Values near 1 are inconclusive for that particular question, even if the test clearly established that the two people are biologically related.
Why Some Results Come Back Inconclusive
An inconclusive result does not mean the lab made a mistake. It means the DNA evidence genuinely does not point strongly in either direction. There are several reasons this happens:
- Overlapping ranges: Half siblings, full siblings, and even first cousins or uncle-nephew pairs can produce similar amounts of shared DNA. When two people fall in the overlap zone, no statistical test can confidently pick a single relationship.
- Too few markers: Older tests that examine fewer genetic locations carry less statistical power. Moving from 15 to 21 markers dramatically increases the share of sibling pairs that produce strong results.1Legal Medicine. Sibling assessment based on likelihood ratio and total number of shared alleles using 21 short tandem repeat loci included in the GlobalFiler™ kit
- No parents tested: When neither parent provides a DNA sample, the lab loses a major anchor. Including even one parent’s DNA typically tightens the analysis and can push a borderline case into conclusive territory.
If your report is inconclusive, the most productive next step is usually to test additional family members, especially a known shared parent or other confirmed relatives. Some labs also offer expanded panels or different marker types that can break a tie.
How Your Ethnic Background Affects the Numbers
Every likelihood ratio depends on allele frequencies, which describe how common each genetic variant is in a reference population. If the lab calculates your LR using frequency data from a population that does not match yours, the result can be skewed. Research on forensic familial searching found that when allele frequencies were misspecified, meaning the reference population did not match the tested individual’s actual ancestry, false-positive rates for identifying relatives increased substantially.3PubMed Central. Human-Genetic Ancestry Inference and False Positives in Forensic Familial Searching
A related study found the problem works in both directions. When individuals’ ancestry was well represented in the frequency database, identification rates were more accurate. But for individuals whose ancestry did not match any of the available reference populations (in that study, Native American and Vietnamese American profiles run against African American, European American, and Latino American frequency databases), the statistical method produced inflated match rates.4PLOS ONE. The Influence of Relatives on the Efficiency and Error Rate of Familial Searching
What this means for you: check whether your report mentions which population database was used. Reputable labs will state something like “calculations performed using Caucasian allele frequencies” or list multiple databases. If your ancestry does not align with the database used, the LR and probability percentage may be less reliable. This does not invalidate the test, but it is a reason to treat borderline results with extra caution, and potentially to request recalculation with a more appropriate reference set if the lab offers that option.
When X-Chromosome or Y-Chromosome Markers Help
Standard sibling tests use autosomal markers, the genetic locations found on the 22 non-sex chromosomes that everyone inherits from both parents. But certain family questions are better addressed by looking at sex-linked chromosomes, and your report might include results from X-chromosome or Y-chromosome markers if the lab judged them relevant.
X-Chromosome Markers
The X chromosome has an unusual inheritance pattern. Fathers pass their single X to all their daughters but none of their sons. This means two sisters who share the same father will always share at least one complete X chromosome from him. That shared inheritance makes X-chromosome markers particularly powerful for confirming relationships between women, especially half-sisters who share a father but have different mothers. One case study found that adding 31 X-chromosome markers to a standard autosomal panel produced overwhelmingly strong evidence for half-sisterhood, leading the researchers to suggest that X markers could be the first choice for identifying half-sisters from different mothers.5Journal of Forensic Science and Medicine. Identification of Half-Sisters from Different Mothers by Autosomal and X Chromosomal Short Tandem Repeats: A Case Study
In another case, autosomal markers alone produced a likelihood ratio too low for legal purposes. Adding just six X-chromosome markers boosted the combined LR enough to reach a probability above 99.9%.6Forensic Science International: Genetics Supplement Series. Paternity testing involving human remains identification and putative half sister: Usefulness of an X-hexaplex STR markers More broadly, X-chromosome markers have been recognized as especially useful in kinship scenarios that autosomal markers struggle with, including half-sibling cases, uncle-niece relationships, and situations where an alleged father is unavailable for testing.7Handbook of DNA Profiling. Usefulness of the X-Chromosome on Forensic Science Software tools used in forensic genetics have confirmed that combining standard autosomal panels with X-chromosome markers significantly improves accuracy for half-sibling and other difficult relationship tests.8Forensic Science International. KinshipLR: software development and application for complex kinship testing in forensic genetics
Y-Chromosome Markers
The Y chromosome passes from father to son virtually unchanged across generations. If two males share a father (or a paternal grandfather, or a great-grandfather), they should carry the same Y-chromosome profile. This makes Y markers useful for a specific question: do two men belong to the same paternal line? If the Y profiles match, that is consistent with a shared father, though it is also consistent with more distant paternal relatives like cousins. If the profiles clearly differ, that strongly excludes a shared paternal line.
The catch is that Y-chromosome data is better for ruling relationships out than for confirming them. Because the Y is inherited as a block within a male line, many men in the same extended family share identical or near-identical Y profiles. Population substructure is a real issue, meaning unrelated men from the same community can sometimes share Y haplotypes by coincidence.9PubMed. The Y chromosome in forensic analysis and paternity testing Mutations do occasionally occur and can complicate interpretation. One case involving suspected paternal half-brothers found three mutations across their Y-chromosome profiles, which initially appeared to exclude a relationship. Only by combining the Y data with autosomal markers were the analysts able to confirm the biological link.10PubMed. Three mutations at a Y-STR haplotype defy a paternal half-brothers kinship case analysis
If your report includes Y-chromosome results and shows a match, treat it as supporting evidence rather than proof of siblingship on its own. If it shows a clear mismatch (more than a mutation or two apart), that is strong evidence against a shared father.
Silent Genes and Other Technical Quirks
Occasionally, the DNA at a tested location fails to amplify properly during the laboratory process. This can happen when a small mutation sits in the region where the testing probe binds, preventing the lab’s equipment from detecting one of a person’s two copies of that gene. The result is called a “silent gene” or a null allele. On your report, it might show up as an apparent mismatch at a single location, or it may be noted in the lab’s technical comments.
Silent genes matter because they can alter the likelihood ratio. The lab’s calculations assume they are seeing both copies of each genetic location. When one copy is invisible, the math can drift. Research into this problem has found that the effect on likelihood ratios depends on whether the lab accounts for the possibility of silent genes in its calculations. Labs that adjust for potential null alleles produce more accurate ratios than those that take the observed data at face value.11Forensic Science International. Evaluating the effects of silent genes on pairwise kinship testing
For most consumer-level tests, silent genes at one or two locations will not dramatically change a strong result. But if your report is borderline and you notice a footnote about a failed marker or an unusual mismatch at a single locus, it is worth asking the lab whether a null allele might be involved and whether the LR was adjusted accordingly.
What the Report Cannot Tell You on Its Own
A sibling DNA test establishes a statistical likelihood of a biological relationship. It does not definitively name which parent is shared, and it does not always distinguish clearly between relationship types that produce similar amounts of shared DNA. Half siblings, uncle-nephew pairs, and grandparent-grandchild pairs can all fall into overlapping statistical zones. If the question is specifically “do we share a father?” versus “do we share a mother?” the autosomal result alone will not answer that. Y-chromosome testing can address paternal-line questions for males, and X-chromosome testing can address certain maternal or paternal patterns for females, but the standard autosomal report is agnostic about which parent is shared.
Context matters more than people realize. If you know you share a mother and are testing to determine whether you share a father, tell the lab. If a mother’s DNA sample is available, provide it. Every piece of known-relationship information the lab can incorporate makes the statistics tighter and the result more reliable. A result that looks inconclusive with just two siblings’ DNA can become definitive once a parent’s sample is added.
When Results Reveal Something You Did Not Expect
Sibling DNA tests sometimes uncover non-paternity events, where the person someone believed was their biological father turns out not to be. The rise of direct-to-consumer DNA testing has made these discoveries far more common, and the psychological fallout can be serious. A study of people who discovered a non-paternity event through DNA testing found elevated levels of depression, anxiety, and panic symptoms compared to people who had not experienced such a discovery. The effect was worse when the discovery damaged the individual’s relationship with or attitude toward their mother.12Psychiatry Research. Discovering your presumed father is not your biological father: Psychiatric ramifications of independently uncovered non-paternity events resulting from direct-to-consumer DNA testing
If your sibling test results suggest a relationship you were not expecting, whether it is a half-sibling result when you assumed full siblingship or an “unrelated” finding for someone you grew up with, it is worth sitting with the data before acting on it. Confirm the result with a second test or an expanded panel before having difficult family conversations. Seek support from a genetic counselor, who can walk you through the report’s nuances and help you think through next steps. The numbers on the page are probabilities, not certainties. What they mean for your family is something no lab report can calculate.