Allele-age inference¶

Estimate the time since a specified allele arose from linked variation among its carriers under an explicit genetic and population model.

Version: v1 · 2026-10-17 · History
Domain-specific #: 14057
Domain group: Natural Sciences
Origin domain: Biology & Ecology
Subdomain: Population Genetics → Biology & Ecology

Core Idea¶

Allele-age inference is an analytical method for estimating how long ago a specified genetic variant arose. It reads changes among chromosomes that carry the variant as traces of inheritance through generations: recombination can break up linked sequence, mutation can alter linked markers, and a genealogical or likelihood model converts the observed pattern into a conditional age. The result is an estimate under stated rates, sampling, and population-history assumptions, not a directly observed mutation date.^[1]^[2]

The target event needs care. The first occurrence of a deletion, the common ancestor of sampled carriers, and the origin or expansion of a shared surrounding haplotype need not be the same event. Stephens and colleagues estimated an approximately 700-year age for an ancestral CCR5-Δ32-bearing haplotype and described the deletion's age as only possibly the same. A later reanalysis obtained much older values after revising the genetic map and adding markers; ancient-genome work subsequently placed the deletion at least 6,700 years before present. Those studies demonstrate why the event and model belong beside every reported number.^[2]^[3]^[4]

Present allele frequency can supply a separate model-dependent estimate or a comparison with linked-variation evidence. It is optional, not an obligatory second clock. Disagreement between frequency and haplotype analyses may motivate investigation of selection, drift, migration, or incorrect calibration; the disagreement alone does not establish any one history.^[1]^[3]

Structural Signature¶

Sig role-phrases:

Target origin event. Specify the allele and the event to date. Calling a carrier-haplotype age the mutation's exact origin without a linking argument changes the target of inference.^[2]
Carrier sample and linked evidence. Observe the variant in carrier chromosomes or families and characterize nearby inherited variation. The CCR5 study used linked markers on carrier chromosomes; the Baltic BRCA1 founder study used haplotypes from unrelated carrier families. A frequency count alone would be a different age-estimation route.^[2]^[5]
Genealogical change model. Relate preserved and altered linked material to elapsed generations using recombination, mutation, coalescence, or founder-likelihood assumptions. Shared sequence by itself indicates relatedness; the model supplies the temporal transformation.^[1]^[2]^[5]
Population and time calibration. Declare the applicable marker rates, genetic map, demographic or selection assumptions, and generation-to-year conversion. The CCR5 reanalysis shows how changing a genetic map can shift an age substantially; the Baltic report explicitly converts 62 generations using 25 years per generation. Frequency may constrain a model, but is not required in every case.^[3]^[5]
Qualified age readout. Report which event the model estimated, its units and assumptions, and an interval or distribution if available. An approximate point estimate remains an estimate: the Baltic original abstract gives roughly 62 generations or 1,550 years without an uncertainty interval in the inspected text.^[2]^[5]

The procedure is target → linked carrier observations → temporal model and calibration → qualified readout. Study design and marker technology can change without removing those roles. Remove the target declaration, linked evidence, or temporal model, and this particular inference procedure is no longer established.

What It Is Not¶

Not the age property itself. “Allele age” names the elapsed time sought; this entry names the procedure for inferring it from carrier-linked evidence.
Not automatically the carrier MRCA or ancestral-haplotype age. Those are possible genealogical proxies. A claim about mutation origin needs the model connecting proxy and event; Stephens expressly allowed the CCR5 deletion to be older than the haplotype estimate.^[2]
Not a mandatory comparison of two clocks. Frequency-only models and linked-variation models can be used separately or compared. This admitted procedure uses the latter; frequency is an optional additional constraint.^[1]
Not a selection test by itself. A high frequency with extended linkage can motivate a selection hypothesis, but maps, genomic controls, demography, and ancient observations can change the interpretation. No dated allele alone identifies a historical pathogen or proves which linked variant was favored.^[3]^[4]
Not a clinical prognosis. The inferred age of a disease-associated variant does not by itself establish penetrance, individual risk, or treatment choice.

Scope of Application¶

In population genetics, carrier haplotypes and linked markers are used to reconstruct the history of a variant in a sampled population. CCR5-Δ32 is an instructive case because a 1998 estimate derived from nearby microsatellite variation was later recalculated with a revised genetic map and denser markers. The literal method is the conversion of linked variation into an age under a population model; the historical claim of medieval origin is not a reusable premise.^[2]^[3]

In medical founder-variant research, unrelated families carrying the same pathogenic variant can be compared for inherited surrounding haplotypes. Janavičius and colleagues reported such an analysis for BRCA1 c.4035delA across four Baltic-region countries and a maximum-likelihood estimate of its age. The inspected original abstract supports that application, but not its detailed marker scheme or likelihood formula.^[5]

Ancient DNA can constrain a modern-carrier estimate by showing that the allele existed in a dated sample, as in the later CCR5 work. That direct observation yields a minimum antiquity for a sampled variant, not the exact time or location of its first occurrence. Outside allele-bearing genetic lineages, the named method does not travel literally just because another field also dates an event.^[4]

Clarity¶

The method separates the unknown event from the evidence used to infer it. A shared carrier haplotype can be dated under a model, yet that date need not equal the deletion's first occurrence. Reporting “CCR5-Δ32 is 700 years old” erases this distinction and contradicts later ancient evidence; reporting the 1998 ancestral-haplotype estimate as a historical model output preserves what the source actually established.^[2]^[4]

It also separates a genetic estimate from an evolutionary explanation. The 1998 authors interpreted a young haplotype and comparatively high frequency as consistent with strong selection. Sabeti and colleagues found that the apparent long-range pattern was not a clear genomic outlier and obtained older ages under a revised map. Ravn and colleagues later reported evidence for selection on the associated haplotype in a much earlier period. These are different evidential claims at different resolutions, not one stable “selection clock.”^[2]^[3]^[4]

Manages Complexity¶

Many local details can affect an age estimate: which carriers were sampled, which flanking markers were typed, how linkage was measured, the mutation and recombination rates, the population model, and the conversion from generations to years. The five structural roles compress those details into a reproducible audit: target, evidence, change model, calibration, and readout. This permits two studies of one allele to be compared without pretending that their numerical outputs are interchangeable.^[1]^[3]

The compression has a limit. The Baltic founder estimate and the CCR5 population-history analyses both fill those roles, but they do so with different samples and inference machinery. The original Baltic abstract does not supply detailed model diagnostics; the entry therefore keeps that case at the level the source warrants. A single date without the model and target event would hide the very information needed to reconcile results.^[5]

Abstract Reasoning¶

First define the allele and ask what the proposed date means: first mutation, carrier coalescence, or expansion of a linked ancestral haplotype. Then specify the sampled carrier chromosomes or families and the linked variation observed. Choose a change model and record the map, rate, and demographic assumptions that turn that variation into generations. Convert to calendar time only with an explicit generation-length assumption, and report uncertainty when the source provides it.^[1]^[2]^[5]

Next vary one assumption at a time. If a revised genetic map yields a much older age from substantially the same CCR5 question, the original point estimate was sensitive to calibration. If a dated ancient carrier predates a proposed mutation-origin date, that origin claim cannot stand, even though the ancient sample does not reveal when the mutation first arose. Those tests support a revised inference without converting a proxy into an exact timestamp.^[3]^[4]

Finally, treat selection as a separate historical inference. Ask whether the observed allele frequency and haplotype pattern are unusual against relevant genomic controls or dated frequency trajectories. An age discrepancy can frame that question; it cannot answer it without an explicit comparison model.^[3]^[4]

Knowledge Transfer¶

The target/evidence/model/calibration/readout sequence transfers literally between population and medical genetics when the evidence still consists of allele-bearing lineages and linked genetic variation. It lets an analyst carry the caution about mutation-origin versus carrier-haplotype age from the CCR5 case to a founder-family study. It does not carry the CCR5 selection debate, its genetic map, or its numerical ages into BRCA1; each population and variant requires its own evidence.^[2]^[5]

Outside genetics, estimating an unobserved event time from partial traces instantiates a broader estimation pattern. Calling a document's creation-date estimate or an archaeological chronology allele-age inference would be metaphorical: those settings lack a specified allele, carrier genealogy, and recombination or mutation-linked evidence. The useful transfer is the broader discipline of declaring target and calibration, not the domain name.^[1]

Examples¶

Canonical: CCR5-Δ32 carrier haplotypes and their reanalysis¶

Stephens and colleagues examined CCR5-Δ32-bearing chromosomes and linked microsatellite haplotypes. A common proposed ancestral haplotype and a model of marker changes led to an estimate near 700 years, with a reported 275–1,875-year range. The authors described it as the age of the deletion-bearing ancestral haplotype and only possibly the deletion itself. Sabeti and colleagues later used an improved genetic map and additional markers to obtain older values; Ravn and colleagues' ancient-genome study placed the deletion at least 6,700 years before present. The case demonstrates both how the inference is performed and how its target and calibration can be challenged.^[2]^[3]^[4]

Mapped back: the target origin event is the deletion's first occurrence, distinguished from the ancestral haplotype's expansion; the carrier sample and linked evidence are the deletion-bearing chromosomes and their marker patterns; the genealogical change model maps preserved and derivative haplotypes to generations; population and time calibration includes recombination mapping, marker-change rates, and demographic assumptions; the qualified age readout is each study's explicitly conditional result rather than one accepted exact date. The later ancient observations bound the plausibility of the historical young-origin interpretation.

Applied comparison: Baltic BRCA1 c.4035delA founder families¶

Janavičius and colleagues genotyped 78 unrelated families carrying BRCA1 c.4035delA in Lithuania, Latvia, Poland, and Russia. Their original abstract reports haplotype analysis and a maximum-likelihood estimate of approximately 62 generations, converted with 25 years per generation to about 1,550 years. It provides a second, founder-family application rather than a claim about CCR5-like selection or frequency-clock disagreement. The full article's detailed markers, likelihood function, and uncertainty interval were not inspected for this entry.^[5]

Mapped back: the target origin event is the specified founder allele's appearance; the carrier sample and linked evidence are those families' haplotypes; the genealogical change model is the reported maximum-likelihood inference from haplotype analysis; population and time calibration includes the stated 25-year generation conversion; the qualified age readout is the paper's approximate 62-generation estimate. The inspected abstract does not permit a more detailed reconstruction of the model.

Structural Tensions¶

T1: Broad modern carrier evidence versus direct ancient time anchors. Modern carrier samples can reveal many linked-haplotype changes but require a genetic map and historical model to translate them into time. Dated ancient carriers anchor presence in an earlier period, yet ancient genomes are scarce, unevenly sampled, degraded, and difficult to genotype at a deletion. Favoring the modern proxy can yield broad comparison while risking calibration error; favoring only direct ancient observations gives a defensible minimum age while leaving earlier unsampled history open. Sabeti's revised CCR5 map and Ravn's haplotype-aware ancient-genome analysis show why the two evidence routes must be reconciled rather than treated as equivalent clocks.^[3]^[4] Diagnostic: Does a dated ancient carrier falsify the proposed origin date, and how much older could the allele still be given sampling gaps?

T2: Age inference versus causal selection story. An age estimate alongside present frequency invites an explanation for how an allele became common. A selection narrative may connect genetic pattern to population history, but the same observations can be sensitive to demography, maps, controls, and the particular linked variant favored. Committing early can make a conditional result appear causal; withholding all explanation can leave a useful historical hypothesis untested. Stephens proposed recent selection, Sabeti found no clear recent genomic outlier, and Ravn reported evidence for older selection on the associated CCR5 haplotype, without isolating the deletion as the selected target.^[2]^[3]^[4] Diagnostic: Which comparison data distinguish selection from alternative histories at the estimated time scale, and do they identify the selected target?

Structural–Framed Character¶

Allele-age inference lies toward the structural end of the structural–framed spectrum: its target, linked observations, model, calibration, and output can be stated as an explicit procedure. Its evaluative weight is limited to whether an age estimate is warranted for a scientific purpose; the method itself does not rank a variant's clinical value. It arose as a research procedure in population and medical genetics, rather than as an institutional rule or public convention. Its carriers and rate data are biological, while sampling and model decisions are human scientific practices. The distinction between mutation origin and carrier genealogy belongs to technical genetic vocabulary.^[1]^[5]

Terms such as origin, clock, and Inference travel broadly, yet importing this name into another field would smuggle in allele-bearing chromosomes and linked genetic change. A new founder variant can be recognized as a literal case by finding those roles; a non-genetic dated event would only resemble the broader estimation structure. Its character: a comparatively stable inferential role pattern, framed by biological carriers and genetic research practice, whose literal membership remains within allele-bearing lineages.

Structural Core vs. Domain Accent¶

The live parent is Analytical Method. Its reusable structure asks a defined question, selects evidence, applies a transformation, states a validity regime, checks quality or limits, and reports a bounded finding. This entry fills those roles with an allele-origin question, carrier haplotypes, a genealogical time model, genetic calibration, model checks and a qualified age result. Analytical Method also covers content or chemical analysis without genetic carriers. The broader Estimation idea describes the numerical output, but its current strict graph ancestry requires a formal error measure or bound and purpose tolerance not established for the abstract-only Baltic case; it is a conceptual neighbor, not an additional edge.

The domain accent is indispensable. This entry requires a specified allele, linked variation among carriers, a genealogy or founder model, and rates or maps connecting genetic change to generations. Removing those leaves a general age estimate, not allele-age inference. The method's model sensitivity and event distinction make sense inside genetics; cross-domain resemblance alone cannot promote this named procedure to a prime.

This entry is a kind of Analytical Method.

Allele-age inference is, in every case, a kind of Analytical Method. Every instance answers a genetic-origin question by selecting carrier evidence, transforming it under stated assumptions, examining quality or limits and reporting a qualified result. Analytical methods also exist that involve no allele at all, so allele-age inference is a narrower kind, set apart by its genetic features. Estimation names a related output pattern, but it is not counted as broader: Estimation is always a kind of Approximation, which requires an error bound and a tolerance that the abstract of the Baltic BRCA1 study does not establish. Inference is broader still; Calibration concerns only a possible supporting operation. Most Recent Common Ancestor is a genealogical neighbor or possible proxy; its dated common ancestor is not automatically the allele's first occurrence.

Relationships to Other Abstractions¶

Current abstraction Allele-age inference Domain-specific

Parents (1) — more general patterns this builds on

Allele-age inference is a kind of Analytical Method Domain-specific

Allele-age inference is an analytical method specialized to dating a specified allele from linked carrier variation under an explicit temporal model.

Hierarchy path (1) — routes to 1 parentless root

Allele-age inference → Analytical Method

Neighborhood in Abstraction Space¶

Allele-age inference sits in a sparse region of the domain-specific corpus (77^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¶

Allele age is the target quantity, whereas this entry is one linked-variation procedure for estimating it. Mutation age can denote that quantity too, but naming it does not establish which event a paper dated. Carrier MRCA is a common ancestor of sampled variant-bearing lineages, and ancestral-haplotype age dates a particular linked background; neither equals the mutation event by definition. Frequency-based age estimation uses a population model and current allele frequency, and may be compared with this method but is not a required component. Selection inference asks why frequencies changed, a further question requiring its own evidence.^[1]^[2]

References¶

[1] Slatkin, Montgomery, and Bruce Rannala. “Estimating Allele Age.” Annual Review of Genomics and Human Genetics 1 (2000): 225–249. https://doi.org/10.1146/annurev.genom.1.1.225. Indexed review abstract: https://pubmed.ncbi.nlm.nih.gov/11701630/. Used for method-family framing, not as an original positive case. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i

[2] Stephens, J. Claiborne, et al. “Dating the Origin of the CCR5-Delta32 AIDS-Resistance Allele by the Coalescence of Haplotypes.” American Journal of Human Genetics 62, no. 6 (1998): 1507–1515. https://doi.org/10.1086/301867. Author-hosted published PDF: https://reich.hms.harvard.edu/sites/reich.hms.harvard.edu/files/inline-files/Stephens.1998_0.pdf. Original population-genetic study; the approximate 700-year figure is for an ancestral deletion-bearing haplotype and only possibly the deletion. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o

[3] Sabeti, Pardis C., et al. “The Case for Selection at CCR5-Delta32.” PLoS Biology 3, no. 11 (2005): e378. https://doi.org/10.1371/journal.pbio.0030378. Author-hosted published PDF: https://reich.hms.harvard.edu/sites/reich.hms.harvard.edu/files/inline-files/Sabeti.2005_0.pdf. Original reanalysis using revised maps, denser markers, and genomic controls; it does not rule out every form of selection. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l

[4] Ravn, Kirstine, et al. “Tracing the Evolutionary History of the CCR5delta32 Deletion via Ancient and Modern Genomes.” Cell 188 (2025): 3679–3695. https://doi.org/10.1016/j.cell.2025.04.015. Published PDF in the Cambridge repository: https://www.repository.cam.ac.uk/bitstreams/5d4c418f-51dc-4559-99c1-3dec9319e83f/download. Original ancient/modern genome study; the reported selection signal concerns the associated haplotype. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j

[5] Janavičius, Ramūnas, et al. “Haplotype Analysis and Ancient Origin of the BRCA1 c.4035delA Baltic Founder Mutation.” European Journal of Medical Genetics 56, no. 3 (2013): 125–130. https://doi.org/10.1016/j.ejmg.2012.12.007. Original indexed abstract: https://pubmed.ncbi.nlm.nih.gov/23274591/. Only the original abstract and metadata were inspected here; detailed methods and uncertainty interval are not claimed. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j