{
  "id": "ancient-dna",
  "title": "Ancient DNA",
  "lang": "en",
  "source_lang": "en",
  "summary": "Ancient DNA is genetic material recovered from historical and prehistoric remains, used to investigate past organisms, populations, diseases, and environments.",
  "category": "science",
  "subcategory": "biology",
  "type": "concept",
  "aliases": [
    "aDNA",
    "ancient genetic material"
  ],
  "generated_at": "2026-11-03",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/ancient-dna/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/ancient-dna/",
      "json": "https://aiwiki.page/zh/ancient-dna.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Ancient DNA is DNA recovered from organisms or biological traces preserved in archaeological, historical, or geological material. Sources include bones, teeth, mummified tissues, hair, preserved plants, dental calculus, coprolites, and sediments. The term describes a molecular condition as well as an age: ancient DNA is typically fragmented, chemically altered, scarce, and mixed with DNA from organisms that colonized a specimen after its deposition. There is no universal age threshold separating ancient from modern DNA.\n\nThe field connects molecular biology with archaeology, paleontology, and environmental history. By directly sampling genetic variation at different times, it can test reconstructions previously based on living organisms alone. Its findings have transformed the study of human origins, population movements, domestication, extinct ecosystems, and infectious disease. Its evidential power nevertheless depends on demonstrating that recovered molecules are genuinely ancient, assigning them to the correct organisms, and interpreting incomplete samples within their archaeological and environmental contexts."
    },
    {
      "heading": "Development of the field",
      "text": "The first widely recognized ancient-DNA study appeared in 1984, when Russell Higuchi and colleagues recovered short mitochondrial sequences from a museum specimen of the quagga, an extinct relative of the plains zebra. The specimen was approximately 140 years old. In 1985, Svante Pääbo reported DNA sequences from an Egyptian mummy. These early investigations established the possibility of obtaining genetic information from preserved remains, while also exposing a central difficulty: trace amounts of modern DNA could be more readily recovered than the damaged molecules under investigation.\n\nThe introduction of the polymerase chain reaction made it possible to amplify selected DNA sequences from very small starting quantities. During the late 1980s and 1990s, researchers used it to investigate extinct animals, archaeological humans, and preserved plants. However, amplification also magnified contamination. Reports of DNA from organisms millions of years old, including material associated with amber and dinosaur fossils, often failed subsequent authentication. These failures helped establish requirements for dedicated laboratories, extraction controls, molecular damage analysis, and, where appropriate, independent replication.\n\nA major milestone came in 1997 with the publication of a mitochondrial sequence from the Neanderthal type specimen found in Germany’s Neander Valley in 1856. It showed that Neanderthal mitochondrial diversity lay outside the variation then documented in living humans. Because mitochondrial DNA represents only one line of inheritance, this result could not by itself resolve whether Neanderthals had contributed other parts of their genomes to modern humans. It nevertheless demonstrated that genetically informative molecules could survive in Pleistocene human remains.\n\nHigh-throughput DNA sequencing changed the scale of the field during the 2000s. Instead of amplifying a few selected fragments, researchers could sequence millions of molecules, including damaged fragments too short for conventional amplification assays. A draft Neanderthal nuclear genome published in 2010 provided evidence of interbreeding with ancestors of present-day non-African populations. That year, genetic analysis of a finger bone from Denisova Cave also revealed an archaic human lineage initially recognized through DNA rather than distinctive skeletal anatomy.\n\nSubsequent improvements extended both geographic coverage and temporal reach. A horse genome reported in 2013 came from material approximately 560,000–780,000 years old. Mammoth DNA published in 2021 exceeded one million years in age. In 2022, researchers reported approximately two-million-year-old environmental DNA from northern Greenland. These examples represent exceptionally favorable preservation, not a routine age range. Pääbo received the 2022 Nobel Prize in Physiology or Medicine for discoveries concerning extinct hominin genomes and human evolution."
    },
    {
      "heading": "Molecular decay and the limits of preservation",
      "text": "DNA preservation begins with processes occurring after death. In living cells, repair systems continually correct molecular damage. Once those systems cease functioning, endogenous enzymes, microbial activity, and spontaneous chemical reactions break down tissues and genetic material. Most surviving ancient molecules are consequently much shorter than DNA extracted from fresh tissue, often measuring only a few tens of base pairs. Their abundance depends on the specimen’s history rather than age alone.\n\nOne important mechanism is hydrolysis, in which reactions involving water alter DNA’s chemical structure. Loss of purine bases can leave unstable sites that promote strand breakage. Cytosine can undergo deamination to form uracil; during sequencing, this frequently appears as a cytosine-to-thymine substitution. Because exposed, single-stranded ends are particularly vulnerable, these substitutions tend to concentrate near fragment termini. Oxidation and cross-linking can further modify molecules or prevent their recovery and copying.\n\nTemperature strongly affects the rates of these reactions. Cold environments, especially permafrost, can greatly prolong molecular survival. Warm and humid environments generally accelerate degradation, helping explain why tropical regions remain relatively difficult to investigate despite their archaeological importance. Stable burial conditions, limited water movement, and protective mineral associations may also improve survival. Repeated freezing and thawing, excavation, washing, chemical treatment, and museum storage can change the condition of a specimen after its original deposition.\n\nDNA may persist partly because it binds to mineral surfaces or becomes enclosed within dense tissues. Teeth and the petrous portion of the temporal bone, which surrounds the inner ear, often preserve relatively high proportions of endogenous human DNA. In sediments, adsorption onto clays and other minerals can protect short fragments from degradation. These processes do not preserve an intact organism: surviving molecules are scattered remnants, usually mixed with much larger amounts of microbial and environmental DNA.\n\nThere is no single molecular expiration date. A frequently cited 2012 study of moa bones estimated a half-life of about 521 years for a particular 242-base-pair mitochondrial fragment under the conditions represented by its sample set. This was not a universal half-life for all DNA or all environments. Different fragment lengths, temperatures, and burial chemistries produce different survival probabilities. The oldest authenticated sequences therefore constrain what is demonstrably possible under exceptional conditions, while offering no basis for expecting usable DNA from much older fossils such as non-avian dinosaurs."
    },
    {
      "heading": "Archaeological context and sampling",
      "text": "The scientific value of ancient DNA depends first on knowing what was sampled. A sequence may identify an organism or reveal ancestry, but its historical meaning requires information about provenance, stratigraphy, associated objects, burial practice, and site formation. A specimen’s position in an archaeological layer does not automatically establish its age: burrowing, erosion, redeposition, and later disturbance can move remains between contexts. Sedimentary DNA presents an additional challenge because molecules may have different depositional histories from visible fossils.\n\nDirect radiocarbon dating can anchor many late Pleistocene and Holocene specimens, generally within a practical range extending to roughly 50,000 years under favorable conditions. Older material requires other chronological approaches and geological constraints. Dietary reservoir effects can complicate radiocarbon ages, particularly where people or animals consumed substantial marine or freshwater resources. DNA analysis, isotope measurements, skeletal study, and archaeological dating therefore provide complementary evidence rather than interchangeable clocks.\n\nSampling usually involves removing and powdering a small amount of bone or tooth, but the amount and location matter. Dense inner-ear bone often yields high endogenous DNA proportions, yet sampling it can destroy anatomically valuable structures. Tooth roots, tooth cementum, or minimally invasive sampling of accessible surfaces may offer alternatives. No tissue is universally optimal: preservation differs among sites, and a sample chosen for genetic yield may be poorly suited to answering a different archaeological question.\n\nOther substrates preserve different biological records. Dental calculus, the mineralized form of dental plaque, can contain oral microorganisms and dietary traces. Coprolites may preserve DNA from the producer, food, parasites, and intestinal microbes. Hair shafts can preserve genetic material with relatively little surface contamination after cleaning, although the balance between nuclear and mitochondrial recovery varies. Historical herbarium specimens provide time-series material for studying plants, pathogens, crop dispersal, and changes in genetic diversity.\n\nSampling design also determines which populations become visible. Large cemeteries, cold-climate sites, and well-funded museum collections are easier to study than poorly preserved or inaccessible regions. Sampling several close relatives from one burial can inadvertently overrepresent a family in population-level analyses. Researchers therefore need to distinguish the number of sequenced individuals from the number of independent demographic observations. Recording unsuccessful extractions and preservation assessments can also clarify whether an apparent absence of a population or organism reflects history or merely inadequate molecular survival."
    },
    {
      "heading": "Extraction, sequencing, and genome recovery",
      "text": "Ancient-DNA extraction is designed to recover short, damaged molecules while minimizing the introduction of modern material. Work before amplification generally takes place in dedicated clean facilities physically separated from laboratories containing abundant amplified DNA. Protective clothing, surface decontamination, ultraviolet treatment of appropriate equipment, and restricted movement of materials reduce contamination. Extraction blanks and library blanks accompany samples to reveal DNA introduced through reagents or handling.\n\nFor mineralized tissues, extraction commonly involves dissolving the mineral component with EDTA and digesting proteins with proteinase K. DNA can then be purified using silica-based methods optimized for short fragments. Protocol choice affects which molecules survive the procedure: some methods lose very short sequences, while others improve their recovery at the cost of additional background material. Extraction does not selectively recover only the target organism. A bone extract may contain predominantly microbial DNA acquired during decomposition.\n\nPurified fragments are converted into sequencing libraries by attaching adapter sequences that allow them to be copied and read by sequencing instruments. Double-stranded library methods are widely used; single-stranded methods can recover molecules that are too damaged or too short for some conventional procedures. Index sequences identify libraries so that multiple samples can be sequenced together. Treatment with uracil-DNA glycosylase can remove uracils produced by cytosine deamination, reducing sequencing errors. Partial treatment can preserve some terminal damage while correcting much of the interior sequence.\n\nTwo broad sequencing strategies are common. Shotgun sequencing reads molecules without first selecting particular genomic targets and can provide an overview of the organisms present. Targeted capture uses complementary probes to enrich specified sequences, such as mitochondrial genomes, pathogen genomes, or panels of human genetic variants. Capture can make poorly preserved samples informative, but it concentrates attention on predetermined regions and can introduce biases associated with probe design and sequence similarity.\n\nThe resulting reads are processed computationally, aligned to reference sequences, and screened for duplicates and low-quality matches. Mitochondrial DNA is often recoverable because each cell can contain many copies, whereas most nuclear regions occur in only two copies per diploid cell. Nuclear genomes nevertheless provide far more independent information about ancestry and population history. Genome coverage describes how extensively and repeatedly positions are represented; a low-coverage genome is not a complete sequence. Ancient genomes are normally reconstructed computationally from many overlapping fragments, not recovered as intact chromosomes, and obtaining sequence data does not imply that viable cells or a clonable organism survive."
    },
    {
      "heading": "Authentication and contamination control",
      "text": "Authentication asks whether the sequenced molecules originated in the ancient material and whether their assignment to an organism is reliable. These are related but distinct questions. An authentically old microbial fragment may not belong to the individual whose bone was sampled, while a human sequence matching the expected species may derive from a modern excavator. No single test establishes authenticity in every substrate, time period, and preservation environment.\n\nFragment length and chemical damage provide important evidence. Ancient reads are often short and show elevated terminal cytosine-to-thymine substitutions, with complementary patterns at opposite ends depending on library construction. Analysts can estimate these patterns across many molecules and compare them with expectations for the sample. However, a short fragment is not necessarily ancient, and damage may be weak in relatively recent or exceptionally preserved material. Conversely, old contamination can itself acquire damage.\n\nHuman remains create particular difficulties because modern human DNA closely resembles the target. Contamination can sometimes be estimated from mitochondrial sequence mixtures or from positions expected to differ between the specimen and potential contaminants. In genetically male individuals, the single X chromosome provides another useful signal: excess apparent variation can indicate contributions from additional people. Such estimates depend on sequencing quality, population variation, and the assumptions of the model used.\n\nLaboratory controls identify some contamination routes but cannot reconstruct every event before sampling. A clean extraction blank does not prove that a bone was uncontaminated during excavation or museum handling. Analysts therefore combine laboratory records, molecular damage, fragment distributions, genetic consistency, and contextual evidence. Independent replication remains valuable for exceptional claims, although large-scale sequencing also permits extensive internal checks that were unavailable in early amplification-based studies.\n\nCorrect taxonomic assignment becomes difficult when reads are short, reference databases are incomplete, or closely related organisms share substantial sequence. Conserved DNA may align equally well to several species, while laboratory reagents can contain microbial DNA. Pathogen identification therefore requires more than a few matching reads: distribution across the genome, organism-specific positions, damage, and plausible archaeological context all matter. Likewise, sedimentary detections require controls for movement between layers and for modern environmental input. Authentication is strongest when several independent lines of evidence converge and when alternative explanations are explicitly tested."
    },
    {
      "heading": "Population-genetic analysis and its limits",
      "text": "Ancient genomes provide dated observations of genetic variation, allowing population genetics to examine change directly rather than reconstructing the past entirely from living populations. Researchers investigate relationships among individuals, changes in allele frequencies, population size, and mixture between previously differentiated groups. These analyses distinguish processes such as genetic drift, gene flow, and natural selection, although the same observed pattern can sometimes fit more than one demographic history.\n\nLow coverage requires special treatment. If only one or two reads cover a position, assigning a conventional diploid genotype may produce false certainty. Some analyses instead use genotype likelihoods, which retain uncertainty about possible genotypes. Others use pseudo-haploid calls, selecting one observed allele per position. The latter can support broad population comparisons but discard information about heterozygosity and require statistical methods suited to that data structure.\n\nPrincipal-component analysis can summarize genetic similarities, and model-based clustering can display shared variation. Neither automatically identifies discrete historical peoples. Formal statistics, including \\(f\\)-statistics and \\(D\\)-statistics, test patterns of shared alleles and departures from specified relationships. Mixture models can estimate ancestry proportions relative to selected comparison groups. Their labels describe analytical proxies: an ancient sampled population may resemble a contributing source without being its exact ancestor, and unsampled populations can alter the interpretation.\n\nAncient DNA also reveals biological relationships within burial groups. Close relatives share characteristic amounts and distributions of genetic material; sufficiently informative data can distinguish parent–child pairs, siblings, and more distant kin. Long runs of homozygosity may indicate parental relatedness or small population size, depending on their lengths and frequency. Genetic kinship does not establish social parenthood, household membership, inheritance rules, or cultural identity without additional evidence.\n\nFurther limitations arise from reference bias and uneven sampling. Short damaged reads carrying an allele different from the reference sequence may align less successfully, distorting estimates. Targeted variant panels can favor variation discovered in particular modern populations. Imputation can infer missing genotypes from reference data, but its reliability depends on coverage and genetic similarity to the reference panel. Tests for selection must also account for migration, because changing ancestry can change allele frequencies without local selection. Reported dates and ancestry proportions are therefore estimates conditional on data and models, not exact measurements of uniquely determined historical events."
    },
    {
      "heading": "Human origins, migration, and social history",
      "text": "Ancient DNA has substantially revised human evolution by establishing that several human lineages exchanged genes. The 2010 Neanderthal draft genome showed that ancestors of present-day populations outside Africa interbred with Neanderthals. Most present-day people with substantial non-African ancestry carry roughly 1–2 percent Neanderthal ancestry, although estimates depend on analytical definitions and population history. Gene flow back into Africa and deeper shared history mean that archaic ancestry cannot be described accurately by a simple division between continents.\n\nDenisovan genomes revealed another component of this reticulate history. Denisovan-related ancestry occurs at varying levels in several Asian and Oceanian populations and is particularly substantial in some Oceanian groups. Genomic differences indicate a more complicated history than one mixture event involving a single uniform Denisovan population. In 2018, researchers reported a roughly 90,000-year-old individual from Denisova Cave whose mother was Neanderthal and whose father was Denisovan, providing direct evidence of a first-generation offspring between these lineages.\n\nFor the Holocene, ancient DNA has helped distinguish the movement of people from the spread of practices without substantial migration. Studies published in 2014 and 2015 showed that European population history involved major contributions from hunter-gatherers, early farmers related to populations of Anatolia, and later groups with ancestry associated with the Pontic–Caspian steppe. The spread of farming into much of Europe involved substantial migration, while later steppe-related movements reshaped ancestry in many regions during the third millennium BCE.\n\nOther regions reveal different trajectories rather than repetitions of a European pattern. Ancient genomes have contributed to studies of the settlement and subsequent population history of the Americas, the peopling of island Southeast Asia and Oceania, and population interactions in Africa. Tropical preservation difficulties and historically uneven research investment mean that geographic coverage remains highly unequal. Newly sampled individuals can therefore alter previously plausible models, particularly where earlier reconstructions depended on a small number of genomes.\n\nAt the local scale, combining genetics with burial evidence, isotope analysis, and material culture can investigate mobility, kinship, and community organization. Genetic sex estimates may be compared with skeletal assessment and funerary treatment, but they do not by themselves establish gender identity or social role. Likewise, genetic ancestry is not equivalent to ethnicity, political affiliation, or language. Proposed links between migrations and language dispersals require independent linguistic and archaeological arguments; DNA can document biological movement without identifying what language an individual spoke."
    },
    {
      "heading": "Extinct organisms, domestication, and past ecosystems",
      "text": "Ancient DNA extends phylogenetics beyond living organisms by adding extinct branches to evolutionary comparisons. Quagga sequences clarified its relationship to zebras, while mammoth, cave-bear, moa, and other genomes have revealed relationships obscured by extinction. Directly sampling extinct populations can expose past diversity that is absent from modern descendants. This matters because living populations often represent only a restricted and historically altered subset of a species’ former range.\n\nTime-series data can distinguish long-term population decline from sudden disappearance. They can reveal changes in genetic diversity, population fragmentation, and replacement before extinction. Such evidence does not independently identify a cause of extinction: climate shifts, habitat change, hunting, disease, and interactions among them require ecological and archaeological evaluation. Small samples and uneven geographic representation can also make a regional disappearance appear to be a species-wide demographic event.\n\nDomestication research uses ancient genomes to compare wild ancestors, early managed populations, and later breeds or crop varieties. Studies of dogs, cattle, pigs, horses, and cultivated plants show that domestication was often followed by repeated mixing with wild or geographically distinct populations. An ancient domestic population need not be the principal ancestor of a modern breed. Horse genomes, for example, have traced the expansion of the lineage ancestral to most modern domestic horses during the late third millennium BCE from the lower Volga–Don region, distinguishing it from earlier horse-management contexts.\n\nEnvironmental DNA recovered from sediments, ice, and other substrates can document organisms even where recognizable fossils are absent. This approach may use broad sequencing or targeted assays and often falls within metagenomics. Approximately two-million-year-old DNA reported in 2022 from the Kap København Formation in Greenland revealed a community containing plants and animals unlike those occupying the region today. The evidence included sequences assigned to mastodon relatives, demonstrating the potential of molecular traces to extend known distributions.\n\nEnvironmental records are not simple censuses. Organisms differ in how much DNA they shed, how far it travels, and how well it binds to sediments. A sequence may arrive through feces, roots, water transport, or decaying tissue; abundance of reads is therefore not directly proportional to abundance of organisms. Failure to detect a species does not prove absence. Interpretation depends on sedimentology, independent dating, reference genomes, and comparison with pollen, macrofossils, and other environmental evidence. Together, these records can reconstruct changing ecosystems more fully than any one archive alone."
    },
    {
      "heading": "Ancient pathogens, microbiomes, and molecular phenotypes",
      "text": "Ancient microbial DNA provides direct evidence of infections and microbial communities that cannot be reconstructed reliably from skeletal lesions alone. Teeth are particularly useful for some blood-borne infections because pathogen DNA may be preserved within tissues associated with the dental pulp. Other substrates, including mummified organs, coprolites, and dental calculus, preserve different microbial signals. Detection must distinguish an organism associated with the person during life from microbes introduced during decomposition or burial.\n\nA prominent example is Yersinia pestis, the bacterium responsible for plague. Genomes recovered from victims of the fourteenth-century Black Death established a direct molecular connection between that pandemic and the pathogen. Older genomes have documented Y. pestis in Eurasian individuals from the Late Neolithic and Bronze Age. Some early lineages lacked genetic features important in the flea-associated transmission cycle of later plague strains. Such findings illuminate pathogen evolution, but a genome alone does not establish every transmission route or the clinical course of an ancient infection.\n\nAncient genomes have also informed research on tuberculosis, leprosy, hepatitis B, smallpox, and other infections. A 2014 study recovered tuberculosis-complex genomes from pre-Columbian individuals in Peru that were closely related to strains associated with seals, supporting a transmission history distinct from a simple European introduction. Viral DNA recovered from historical remains has revealed now-extinct lineages and long-term changes in viral populations. RNA generally presents additional preservation difficulties, so the ancient molecular record is particularly uneven among different pathogen groups.\n\nDental calculus can preserve components of the oral microbiome, together with host molecules and traces of consumed or inhaled material. Comparing such samples across periods can investigate changes associated with diet, subsistence, and oral ecology. Nevertheless, a preserved microbial community is filtered by mineralization, molecular survival, and laboratory recovery. It is not an unaltered sample of all organisms present during life, and detecting a food-associated sequence does not necessarily demonstrate routine consumption.\n\nResearchers also investigate molecular features beyond ordinary DNA sequence. Patterns of cytosine damage can sometimes support reconstruction of aspects of ancient DNA methylation, because methylated and unmethylated cytosines have different deamination products. These inferences depend on tissue type, preservation, and reference comparisons. Likewise, predicting appearance or disease susceptibility from ancient variants is limited by incomplete genomes, environmental influences, and the applicability of present-day association studies. Detecting a variant linked to a trait does not establish that a particular individual expressed that trait or experienced a particular disease."
    },
    {
      "heading": "Ethics, stewardship, and access",
      "text": "Ancient-DNA research on human remains involves destructive analysis of people who cannot consent and whose remains may carry continuing cultural, spiritual, or familial significance. Legal permission to sample is not identical to agreement from all relevant communities. Research governance therefore involves museums, archaeological authorities, descendant groups, Indigenous peoples, and local researchers, with responsibilities varying according to the remains, jurisdiction, and historical circumstances.\n\nImportant questions include which remains should be sampled, how much material should be removed, who defines the research objectives, and what happens to unused material. Consultation before sampling can identify restrictions or priorities that would otherwise be overlooked. Alternatives such as sampling already damaged elements, sharing extracts between projects, and screening preservation before further destruction can reduce repeated consumption of finite collections. Human remains also retain scientific value for anatomy, pathology, dating, and future methods not yet developed.\n\nGenomic data raise concerns extending beyond the physical specimen. Ancient individuals can be biologically related to living populations, and findings about ancestry or relatedness can acquire meanings not intended by the original study. Claims concerning group identity, territorial rights, or cultural continuity require particular care because genetic relationships do not automatically determine legal or social membership. Genomic sex and kinship results may also conflict with expectations derived from oral histories or previous archaeological interpretation.\n\nData sharing supports reproducibility by allowing other researchers to reassess read alignments, contamination estimates, and demographic models. At the same time, unrestricted release may conflict with community agreements or principles of data sovereignty. Metadata can reveal sensitive burial locations or information about culturally restricted collections. Access arrangements therefore need to consider both scientific transparency and legitimate stewardship requirements rather than assuming that one policy is suitable for every dataset.\n\nThe global organization of the field also affects interpretation. Sequencing infrastructure, funding, and publication opportunities remain concentrated in a limited number of institutions, while many sampled remains originate elsewhere. Collaborative planning, local analytical capacity, appropriate recognition, and access to results can change who participates in producing historical knowledge. Reporting must distinguish genetic findings from claims about race, ethnicity, nationality, or cultural worth. An ancestry component is a statistical description relative to selected reference samples, not a biological definition of a people, and population histories frequently involve mixture, continuity, and change at several different scales."
    }
  ],
  "links": [
    "dna",
    "molecular-biology",
    "archaeology",
    "paleontology",
    "polymerase-chain-reaction",
    "dna-sequencing",
    "enzyme",
    "hydrolysis",
    "temperature",
    "permafrost",
    "radiocarbon-dating",
    "mitochondrion",
    "population-genetics",
    "allele",
    "genetic-drift",
    "gene-flow",
    "natural-selection",
    "human-evolution",
    "isotope",
    "phylogenetics",
    "environmental-dna",
    "metagenomics",
    "rna",
    "microbiome",
    "dna-methylation"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}