DNA paternity testing¶

A genetic assessment that interprets DNA-marker evidence through inheritance and competing paternal hypotheses to report a supported, excluded or unresolved fatherhood relationship.

Version: v1 · 2026-10-17 · History
Domain-specific #: 14128
Domain group: Natural Sciences
Origin domain: Biology & Ecology
Subdomains: Genetic Parentage Testing, Forensic Genetics → Biology & Ecology
Aliases: DNA paternity test, Genetic paternity testing

Core Idea¶

DNA paternity testing assesses a specified biological-fatherhood hypothesis by interpreting genetic evidence through parental inheritance and relevant paternal alternatives. The evidence may come from directly typed people, fetal genetic contribution inferred from maternal plasma, or relatives who provide information about an unavailable alleged father. The completed assessment connects those observations to a qualified finding: exclusion, support, nonexclusion or an inconclusive result.^[1]^[2]^[3]

The common operation is more specific than collecting DNA and more general than one laboratory algorithm. Genetic markers must bear on the proposed paternal relationship through an inheritance model. A profile alone does not supply that interpretation, and a compatible profile does not remove the need to consider plausible alternatives. Neither a mother–child–alleged-father trio, a direct alleged-father specimen, a particular marker panel nor a numerical posterior is necessary in every instance.^[1]^[2]^[3]

A paternity index expresses evidence under competing hypotheses. It must not be silently treated as a probability that the alleged father is the biological father: a posterior probability also depends on a prior. A laboratory's decision convention is a further matter. The inspected studies use different local combined paternity index (CPI) cutoffs; those numbers do not define the biological relationship or the method as a whole.^[4]^[1]^[3]

Structural Signature¶

The assessment has six connected roles:

Specified fatherhood question: identify the child or fetus and the paternal relationship being assessed. An alleged father can be represented by a hypothesis informed by relatives when his own sample is unavailable. Without the particular paternal question, the same data might support another kinship inquiry instead.^[1]^[2]^[3]
Genetic evidence and sample configuration: establish which genotypes or genetic signals are observed and which must be inferred. Direct child and alleged-father profiles, a maternal-plasma mixture, and a relative configuration supply different evidence. They cannot be treated as interchangeable sample sets.^[1]^[2]^[3]
Marker-to-inheritance interpretation: relate marker observations to parental transmission possibilities. A direct STR comparison, a fetal-mixture SNP model and pedigree-constrained reconstruction implement this role differently. Removing that connection leaves marker data rather than a fatherhood assessment.^[1]^[2]^[3]
Declared paternal alternatives and assumptions: state what the evidence is being compared against and which population, transmission or kinship assumptions support the comparison. An unrelated alternative and a close paternal relative need not receive the same evidential treatment. A numerical CPI is one implementation of this comparison, not a compulsory output of every possible assessment.^[1]^[2]^[3]
Qualified parentage finding: connect the interpreted evidence to the paternal question. The finding can be supportive, exclusionary or unresolved. An assay's reported finding is distinct from a separate study comparing its findings with another method.^[1]^[2]^[3]
Case-specific adequacy and uncertainty boundary: retain the conditions that make the finding informative, including marker coverage, mixture information and available or correctly described relatives. This is an obligation of defensible interpretation, not a claim that every performed assay is valid or that every case resolves.^[1]^[2]^[3]

A conclusive result is not a membership test for the method. An inconclusive investigation can instantiate all six roles while showing that the available evidence does not warrant the desired conclusion. Conversely, a numerical score detached from a paternal hypothesis and its assumptions is not the complete assessment.

What It Is Not¶

It is not DNA profiling alone. Producing a marker profile supplies evidence. Interpreting its implications for a specified paternal relationship supplies the additional operation at issue here.^[1]^[2]

It is not a mandatory trio procedure. Gao and colleagues studied motherless alleged-father–child duos as well as trios. Their fatherless alleged-mother–child duos concern maternity; these must not be confused with paternity investigations in which an unavailable alleged father's genetic information is reconstructed from relatives.^[1]^[3]

It is not a guarantee from a high CPI. Close-relative alternatives can make an apparently strong result misleading. Nor does falling below a local confirmation cutoff automatically establish biological nonpaternity: an inconclusive category can remain between locally defined outcomes.^[1]

It is not all genetic genealogy or kinship testing. Standalone maternity and grandparentage ask different relationship questions. Grandparents or siblings can provide evidence within a paternity assessment without making every grandparentage or sibling test a paternity test. A paternal-line marker match can also fail to distinguish closely related candidate fathers.^[1]^[3]

It is not a legal or social definition of fatherhood. A genetic evidential finding and the institutional consequences assigned to it are different matters. No universal rule about admissibility, consent procedure, custody of samples or legal burden of proof follows from the genetic operation.

Scope of Application¶

The directly inspected positive settings are human parentage investigations. They include postnatal STR comparisons with directly sampled people; prenatal assessment using fetal genetic contribution inferred from maternal plasma; and investigations of a specified deceased or missing alleged father using available relatives. These are unlike whole settings because their observation configurations and latent-genotype problems differ, while their paternal-hypothesis operation remains.^[1]^[2]^[3]

Invasive prenatal material can have a separate role. In Jiang's study, amniotic-fluid evidence supplied a comparison for evaluating the maternal-plasma method. That comparator is not a constitutive input of the plasma assay merely because the researchers used it in the same study. The inspected work does not establish a chorionic-villus paternity case or first-trimester clinical validation.^[2]

The sources do not justify carrying their local reporting thresholds, marker performance or population assumptions into every application. They also do not establish an animal-breeding example, detailed historical priority, special chimerism explanations or country-by-country legal rules. Those subjects need their own evidence rather than inheriting authority from the three human studies.

Clarity¶

The method separates what was observed, what was inferred and what was concluded. A directly measured STR profile, an inferred fetal contribution and an unavailable father's reconstructed genetic information occupy different evidential positions. Calling them all “the father's DNA” can hide which parts of the assessment depend on a model or declared pedigree.^[2]^[3]

It also separates the test finding, validation agreement and posterior or institutional interpretation. A study's agreement with another method bears on assessment of that implementation; it is not itself the paternity finding for a family. A CPI is hypothesis-conditioned evidence; a posterior or a locally accepted report adds further assumptions or conventions. These distinctions prevent a precise number from answering a different question than the one the evidence actually supports.^[2]^[4]

Manages Complexity¶

Many marker values and family relationships can be organized around a smaller set of questions: which paternal hypothesis is at issue, what genetic observations are available, how inheritance connects those observations to the hypothesis, which alternatives matter, and what information deficit limits the finding. The six roles preserve that chain instead of treating every matching allele or extra marker as independent reassurance.

The compression has limits. More loci can increase information, but a sparse or wrongly described family configuration can remain problematic. In a prenatal mixture, fetal fraction and effective sequencing information change what the signal can discriminate. These are different sources of uncertainty and may call for different evidence, rather than a larger number attached to the same unsupported interpretation.^[2]^[3]

Abstract Reasoning¶

A useful inference begins with the form of the uncertainty. If a compatible result could also arise under a close-relative alternative, ask which additional comparison would discriminate those paternal hypotheses. Gao's uncle case illustrates why compatibility and a score calculated under a limited alternative do not automatically settle the particular fatherhood claim.^[1]

If the alleged father is unavailable, the next question is what genetic information his relatives actually constrain. Adding markers to the same small relative set and adding a different informative relative are not automatically equivalent interventions. An unresolved report can be a warranted endpoint when neither supplies sufficient evidence.^[3]

If a plasma result rests on a fetal-mixture model, assess the relevant mixture and sequencing conditions before extending the result to another setting. Jiang's simulated low-fetal-fraction analyses do not substitute for clinical observations in the first trimester. The inference must retain the distinction between a modeled possibility and an observed validation case.^[2]

Knowledge Transfer¶

Transfer between panels and populations requires a model check. The same parentage question can be investigated with different markers, but a study's frequencies, informative-marker coverage and decision convention do not transfer automatically. Gao's kit comparison demonstrates differing outcomes within one study, rather than a universal marker count.^[1]

Transfer from direct samples to prenatal plasma requires a different observation model. The paternal question persists; the evidence is now a mixture from which fetal information is inferred. Fetal fraction, marker informativeness and sequencing conditions become relevant. This is literal reuse of the genetic parentage operation with additional conditions, not evidence that every direct-sample procedure works unchanged on plasma.^[2]

Transfer to an unavailable alleged father requires supported kinship information. Relatives can constrain the missing genetic information, but their identities, relationships and informativeness matter. The ability to begin such an investigation is not a promise that it will conclude.^[3]

Transfer from evidence to a probability or institutional decision requires an explicit bridge. A paternity index, a posterior probability and a laboratory or legal acceptance rule are different objects. Retaining the hypotheses, prior where relevant, and local decision context prevents one study's reporting convention from becoming a universal claim.^[4]^[1]^[3]

Beyond genetic parentage, comparison and evidence reasoning can recur. That does not make a legal analogy, a documentary family argument or another diagnostic task an application of the entire DNA paternity method.

Examples¶

Canonical: direct postnatal STR investigation¶

Gao and colleagues examined 1,442 real alleged-father–child–mother trios and 803 real duos. Of the real duos, 412 were motherless alleged-father–child paternity cases; 391 were fatherless alleged-mother–child maternity cases. The study also calculated duo forms of the trio data; those reanalyses are not additional independently recruited families. Four autosomal panels were compared under Chinese specifications and population-frequency assumptions. Some smaller-panel duo assessments remained inconclusive under the local rule.^[1]

Mapped back: the alleged-father–child relationship supplies the specified fatherhood question. The direct personal STR profiles supply genetic evidence and sample configuration. Their transmission-based comparison supplies marker-to-inheritance interpretation. The statistical assumptions and relevant close-relative possibilities supply declared paternal alternatives and assumptions. The local supportive or inconclusive assessment supplies the qualified parentage finding. Marker coverage and the relative-alternative problem supply the case-specific adequacy and uncertainty boundary.^[1]

The study's CPI cutoff of 10,000 is local. Its close-relative cases caution against equating a large score with a correct conclusion about a particular parent. The maternity cases help clarify the boundary; they are not a second paternity setting.^[1]

Applied: prenatal maternal-plasma SNP investigation¶

Jiang and colleagues developed a SNP-based maternal-plasma paternity assessment in one family and evaluated it in sixteen further pregnancies. The test interpreted mixed plasma DNA using parental information and a fetal/paternal contribution model. It compared evidence under alleged-father and random-man hypotheses, producing plasma-based paternity findings.^[2]

Mapped back: the fetal paternal relationship supplies the specified fatherhood question. Maternal-plasma DNA and available parental genotypes supply genetic evidence and sample configuration. The SNP-mixture model supplies marker-to-inheritance interpretation. Its alternative paternal hypotheses supply declared paternal alternatives and assumptions. The plasma-based interpretation supplies the qualified parentage finding. Fetal fraction, informative loci, sequencing conditions and population scope supply the case-specific adequacy and uncertainty boundary.^[2]

Separate validation evidence: the seventeen plasma findings agreed with seventeen amniotic-fluid sequencing findings. Only five of those families also had a conventional capillary-electrophoresis STR comparison, which agreed in five of five cases. These are distinct comparisons, not seventeen conventional STR validations or a universal accuracy estimate. First-trimester performance was simulated. The paper also gives discordant timing descriptions—13–30+6 weeks in Methods and 12–20 weeks in its Discussion—so one exact recruitment range should not be asserted without qualification.^[2]

Applied: an unavailable alleged father¶

Basgalupp and colleagues retrospectively examined 858 paternity investigations in southern Brazil in which the alleged father was deceased or missing. Genetic profiles from available relatives constrained information about the unavailable father, alongside the proband child and mother. With the study's 17-locus analysis and local reporting criterion, 695 reports were classed as conclusive and 163 remained inconclusive.^[3]

Mapped back: the relationship between a particular child and the unavailable alleged father supplies the specified fatherhood question. Child, maternal and relative profiles supply genetic evidence and sample configuration. Pedigree-constrained reconstruction supplies marker-to-inheritance interpretation. The declared kinship configuration and paternity calculation supply declared paternal alternatives and assumptions. The local conclusive or inconclusive report supplies the qualified parentage finding. Sparse relatives, potentially incorrect reported kinship and marker informativeness supply the case-specific adequacy and uncertainty boundary.^[3]

The reported resolution rate is not an independently established accuracy rate. The local CPI criterion of 1,000 and its probability gloss should not be generalized. Additional markers did not resolve every case with few available relatives; nor does the presence of grandparents guarantee a conclusion in a different investigation.^[3]

Structural Tensions¶

T1: additional marker evidence versus additional relative evidence when the alleged father is unavailable. Deepening marker analysis can extract more information from people already sampled, without depending on a new claimed kin relationship. Obtaining another informative relative can instead add an inheritance constraint that more loci in the same sparse family configuration do not provide. These approaches address different deficits. Favoring additional loci can still leave the paternal information underdetermined; favoring relatives depends on their availability and on the truth of the declared kinship. Basgalupp's findings support this local choice problem, not a universal cost-minimizing strategy or guarantee of resolution.^[3]

Diagnostic: Is the remaining uncertainty principally marker discrimination within a supported pedigree, or an underinformative or unreliable relative configuration?

Structural–Framed Character¶

DNA paternity testing is a structural genetic-evidence method implemented through human-designed practices. Its conclusions are constrained by inheritance, observed markers and the stated alternatives, rather than created by the report. Its evaluative weight must nevertheless be separated from its application: a genetic assessment is not moral approval, legal parenthood or a decision about family obligations.

Human practice shapes the assay, sample selection, model and report. Institutional settings also shape decision conventions, as the different local CPI rules illustrate; no single convention defines every instance. Vocabulary travels unevenly. Evidence, comparison and hypothesis are broadly usable, whereas genotype, paternal transmission, fetal mixture and pedigree retain their biological meanings. Practitioners import instruments and inferential procedures, while using them to assess a relationship that is not constituted by the test itself.^[1]^[2]^[3]

Its character: a domain-bound inheritance-based evidential procedure, structurally constrained in its genetic reasoning and conditionally framed in how its findings are reported and used.

Structural Core vs. Domain Accent¶

Skeletal relation. Observations are interpreted against declared alternatives and conditions to produce a qualified relational finding. The catalog's Comparison names the more portable operation of bringing items into a shared frame and reading off a relation. Recognizing that operation does not reproduce the whole parentage assessment.

Domain-bound accent. Biological paternal transmission, DNA markers, directly observed or latent genotypes, fetal mixtures and pedigree configurations determine which comparisons can bear on the fatherhood hypothesis. They are not interchangeable decorations around an otherwise complete generic test.^[1]^[2]^[3]

Why this is not a Prime. Removing those commitments leaves general comparison or evidence reasoning and loses the named genetic procedure. Variation among human parentage settings establishes reuse within the domain; it does not demonstrate intact cross-substrate transfer. A hypothetical comparison outside genetics would be an analogy unless it supplied a different, independently justified abstraction.

This entry is a kind of Analytical Method.

Analytical Method is the strict genus of this assessment. The paternal question, selected genetic inputs, inheritance-based interpretation, declared assumptions, adequacy checks and qualified finding realize its complete procedure-to-finding structure. DNA paternity testing specializes that structure through genetic transmission and the fatherhood relation; it requires neither a numerical posterior nor a conclusive result in every case.

Comparison is visible in relating genetic observations to the patterns and alternatives under which they acquire significance. Statistical Inference provides a broader vocabulary for conditional evidence and uncertainty. These conceptual relations do not erase the additional biological mechanism or establish that every paternity implementation inherits every requirement of a particular statistical procedure.

Identity Test asks whether presentations refer to the same entity. Paternity instead concerns a relation between different people. The shared word “test” does not make those identities equivalent. Likewise, the presence of competing hypotheses does not by itself turn a likelihood-ratio assessment into a frequentist null-hypothesis significance test.

Relationships to Other Abstractions¶

Current abstraction DNA paternity testing Domain-specific

Parents (1) — more general patterns this builds on

DNA paternity testing is a kind of Analytical Method Domain-specific

DNA paternity testing is an analytical method specialized to inheritance-conditioned genetic evidence and a specified biological-fatherhood question.

Hierarchy path (1) — routes to 1 parentless root

DNA paternity testing → Analytical Method

Neighborhood in Abstraction Space¶

DNA paternity testing sits in a sparse region of the domain-specific corpus (98^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (2721 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-17

Not to Be Confused With¶

Genealogical DNA test. Tell: Is the question a specific paternal relationship, or broader ancestry and genealogy involving databases and documentary evidence?
DNA profile or marker panel. Tell: Are genetic observations merely produced, or interpreted as evidence about the paternal hypothesis?
Maternity or standalone grandparentage assessment. Tell: Which relationship is the finding actually about? Relatives can instead be inputs to a paternity case.
Paternal-line compatibility. Tell: Does the evidence distinguish the particular alleged father from relevant male relatives, or only support a shared lineage?
Posterior probability or local acceptance category. Tell: Which hypotheses, prior and decision convention connect the evidence to that reported number or category?
Legal or social fatherhood. Tell: Is the claim genetic, or an institutional relation with its own criteria and consequences?

References¶

[1] Gao, H., et al. (2019). “Application of CPI cutoff value based on parentage testing of duos and trios typed by four autosomal kits.” PLOS ONE 14(11), e0225174. https://doi.org/10.1371/journal.pone.0225174. Original full text, Materials and methods; Results, Table 1 and close-relative cases; Discussion. The cutoff, panel comparisons and outcomes are study-bounded. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t ↩u ↩v

[2] Jiang, H., et al. (2016). “Noninvasive Prenatal Paternity Testing (NIPAT) through Maternal Plasma DNA Sequencing: A Pilot Study.” PLOS ONE 11(9), e0159385. https://doi.org/10.1371/journal.pone.0159385. Original full text, Methods, including the mathematical model; Results, Table 1 and comparison paragraph; Discussion/limitations. Plasma findings, amniotic-sequencing agreement and the five-case conventional CE-STR comparison are distinct evidence roles. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t ↩u

[3] Basgalupp, S. P., et al. (2014). “Investigation of paternity with alleged father deceased or missing: Analysis of success at the end of the report.” Forensic Science International: Genetics 12, 120–121. https://doi.org/10.1016/j.fsigen.2014.05.010. Institutional full article, printed 120–121, PDF indices 0–1; Methods, Results/discussion and Table 1. Conclusions and local reporting criteria are not universal performance estimates. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t ↩u ↩v ↩w

[4] Gjertson, D. W., et al. (2007). “ISFG: Recommendations on biostatistics in paternity testing.” Forensic Science International: Genetics 1(3–4), 223–231. https://doi.org/10.1016/j.fsigen.2007.06.006. Original abstract and commission publication record. Access used here is the original abstract and indexed commission text; the broad likelihood-ratio/prior distinction is retained, while detailed reporting rules, quotations and equations are not drawn from uninspected full text. registry ↩a ↩b ↩c