henhaobio
Sign in

NatureGeneticsPeer-reviewed journal article

The Genomic History Of Southeastern Europe

Iain MathiesoniD, Songül Alpaslan-RoodenbergiD, Cosimo PosthiD, Anna Szécsényi‐NagyiD, Nadin RohlandiD, Swapan MallickiD, Íñigo OlaldeiD, Nasreen BroomandkhoshbachtiD, Francesca CandilioiD, Olivia CheronetiD, Daniel FernandesiD, Matthew FerryiD, Beatriz GamarraiD, Gloria González FortesiD, Wolfgang HaakiD, Éadaoin HarneyiD, Eppie R. JonesiD, Denise KeatingiD, Ben Krause‐KyoraiD, İşil KüçükkalıpçıiD, Megan MicheliD, Alissa MittnikiD, Kathrin NägeleiD, Mario NovakiD, Jonas OppenheimeriD, , Cristian ViragiD

PDFDOIPreprintCode
In-depth full-text digest available · Jump to digest ↓

Abstract

Farming was first introduced to southeastern Europe in the mid-7th millennium BCE - brought by migrants from Anatolia who settled in the region before spreading throughout Europe. To clarify the dynamics of the interaction between the first farmers and indigenous hunter-gatherers where they first met, we analyze genome-wide ancient DNA data from 223 individuals who lived in southeastern Europe and surrounding regions between 12,000 and 500 BCE. We document previously uncharacterized genetic structure, showing a West-East cline of ancestry in hunter-gatherers, and show that some Aegean farmers had ancestry from a different lineage than the northwestern Anatolian lineage that formed the overwhelming ancestry of other European farmers. We show that the first farmers of northern and western Europe passed through southeastern Europe with limited admixture with local hunter-gatherers, but that some groups mixed extensively, with relatively sex-balanced admixture compared to the male-biased hunter-gatherer admixture that prevailed later in the North and West. Southeastern Europe continued to be a nexus between East and West after farming arrived, with intermittent genetic contact from the Steppe up to 2,000 years before the migration that replaced much of northern Europes population.

In-depth summary

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 4 (19 Sept 2017) · CC BY 4.0

Ancient DNA from 223 people (12,000-500 BCE) reveals a west-east hunter-gatherer cline, largely unadmixed farmers, Caucasus-shifted Aegean farmers, and Balkan steppe ancestry almost 2,000 years early.

Southeastern Europe is where farmers of Anatolian descent first met native foragers, yet earlier ancient DNA from Europe's first farmers came mostly from central and western Europe, with only three southeastern individuals reported. This dataset fills that gap and resolves forager structure, farmer-forager mixing and the timing of steppe contact in the region.

Key results

  1. ShownThe farmers who brought agriculture to northern and western Europe crossed the Balkans with little local forager admixture. Some communities, such as Malak Preslavets, mixed substantially with hunter-gatherers.

    Balkan Neolithic modeled as 98% (95% CI 97-100%) NW Anatolian Neolithic-related ancestry. 8 of 9 Malak Preslavets individuals carry excess forager ancestry, fit by 82% (CI 77-86%) NW Anatolian, 15% (CI 12-17%) western hunter-gatherer (WHG) and 4% (CI 0-9%) eastern hunter-gatherer (EHG) ancestry. Four Copper Age Trypillian individuals are ~80% NW Anatolian-related. · Population transformations in the first farmers

  2. ShownHunter-gatherer admixture into early farmers was sex-balanced. The later Copper Age/Middle Neolithic resurgence of forager ancestry in central Europe and Iberia was male-biased.

    Autosome vs X-chromosome qpAdm Z-scores (positive = male forager bias): Balkan Neolithic 0.27, LBK (Linearbandkeramik) -0.22, Iberian Early Neolithic 0.74; Copper Age Balkans 1.66, Iberia 3.08, Central Europe 2.74. Forager-associated Y haplogroups in 7/9 Iberian Neolithic/Copper Age and 9/10 Middle-Late Neolithic central European males. · Sex-biased admixture between hunter-gatherers…; Population genetic analysis

  3. ShownHunter-gatherers formed a west-east ancestry cline correlated with geography, though an EEMS migration surface suggests clustered rather than smoothly clinal structure. Ancestry shifted over time within regions, and the Iron Gates population may carry ancestry found in neither WHG nor EHG.

    114 of 214 new individuals have almost entirely forager ancestry. Ukraine Neolithic shifts toward WHG vs Mesolithic, D(Mbuti, WHG, Ukraine_Mesolithic, Ukraine_Neolithic) Z=8.9; Latvia shifts the opposite way (Middle Neolithic 65% EHG). Iron Gates (40 individuals) fit ~85% WHG/~15% EHG, but qpAdm p=0.0003; mtDNA K1 in 7/40. · Hunter-Gatherer substructure and transitions; Population genetic analysis; Figure legends

  4. ShownSteppe-related ancestry reached the eastern Balkans sporadically by the mid-5th millennium BCE. Balkan and Anatolian data give no evidence for a large steppe population movement through the Balkans into Anatolia.

    ANI163 (Varna I, 4711-4550 BCE) and I2181 (Smyadovo, 4550-4450 BCE) carry steppe ancestry, almost 2,000 years before earlier evidence this far west; other Copper Age (~5000-4000 BCE) Balkan individuals lack it. Bronze Age (~3400-1100 BCE) Balkans: 30% (CI 26-35%). Bronze Age Anatolians lack both EHG and WHG ancestry. · Population transformations in the first farmers; No evidence that steppe-related ancestry moved…

  5. SuggestedSome southern Greek Neolithic farmers descend from a lineage distinct from the NW Anatolian source of other European farmers, shifted toward Caucasus hunter-gatherer (CHG) ancestry.

    D-statistics place five Peloponnese_Neolithic individuals (~4000 BCE; Diros and Franchthi caves) and one Krepost, Bulgaria individual (I0679_d, 5718-5626 BCE) away from WHG and toward CHG, relative to Anatolian and Balkan Neolithic. By contrast, Iberian Epicardial, LBK and Balkan Neolithic form a clade with NW Anatolia as the outgroup. · A novel source of ancestry in Neolithic Europe

Methods, data and reproducibility

Ancient skeletal DNA was enriched for 1,237,207 SNPs by in-solution capture, called pseudo-haploid and merged with published ancient and present-day panels. The data were analyzed with projected PCA, ADMIXTURE, D-statistics, qpAdm and qpGraph. Autosome-versus-X qpAdm contrasts tested for sex-biased admixture, and EEMS mapped forager migration surfaces.

  • Newly generated 1240k-capture genome-wide data from ancient individuals in the Balkans, Carpathian Basin, North Pontic Steppe and nearby regions, 12,000-500 BCE · human · n = 223 (214 newly reported; 214 in analysis set)
  • Previously published ancient individuals merged at the 1240k sites · human · n = 274
  • Present-day West Eurasians genotyped on the Illumina Human Origins array (HO merge; 597,573 sites) · human · n = 799
  • Simons Genome Diversity Project high-coverage genomes spanning 142 worldwide populations (SGDP merge, used for qpAdm) · human · n = 300
  • Iron Gates hunter-gatherers (Romania/Serbia border) from five sites · human · n = 40
  • Hunter-gatherers dated before ~5000 BCE with no NW Anatolian ancestry, used to fit the EEMS migration surface · human · n = 116
  • New direct AMS radiocarbon bone dates · human · n = 113 dates for 112 individuals
  • Stable carbon and nitrogen isotope values on bone collagen · human · n = 94 samples
Methods & tools
1240k in-solution SNP capture (1,237,207 SNPs), Partial (half) UDG library treatment, Illumina NextSeq 500 / HiSeq 4000 sequencing, bwa (alignment to b37/hg19), smartpca v15100 (lsqproject projection), ADMIXTURE v1.3.0 (supervised and unsupervised), qpDstat v710 (D-statistics), qpAdm v610 (autosomal and X-chromosome ancestry), qpGraph v6021 (AdmixTools), EEMS with bed2diffs2, OxCal 4.2.3 with IntCal13 calibration, binom R package (Agresti-Coull intervals)
Software & environment
smartpca v15100, ADMIXTURE v1.3.0, qpDstat v710, qpAdm v610, qpGraph v6021, OxCal 4.2.3 (IntCal13), ISOGG Y nomenclature version 11.110; reads aligned to b37/hg19 with bwa.
Reproducibility gaps
The text gives no accession or repository for the new sequence data or for analysis scripts. The listed URLs point only to third-party software, and per-sample details are in supplementary tables not included here.
Limitations (6)
  • A two-source WHG plus EHG qpAdm model fits Iron Gates hunter-gatherers poorly (p=0.0003). Their extra ancestry source (Anatolian contact, or a post-glacial southeastern population ancestral to WHG) remains unresolved. · stated by the authors
  • X-chromosome ancestry estimates have large standard errors, and qpAdm can depend on which outgroup is listed first. Sex-bias conclusions therefore needed robustness checks using cyclic permutation of the outgroups. · stated by the authors
  • EEMS output depends on sampling density, and fine-scale features may be uninterpretable. The authors treat only the north-south barrier between WHG- and EHG-dominant groups as robust. · stated by the authors
  • No steppe ancestry in Bronze Age Anatolians does not rule out Indo-European languages reaching Anatolia via the Balkans without large-scale migration. Links to present-day populations also remain largely unresolved. · stated by the authors
  • Several headline inferences rest on few individuals: five Peloponnese Neolithic people for the CHG-shifted lineage, two Copper Age individuals for early steppe ancestry, and Y-chromosome counts of 7/9 and 9/10 for male bias. · noted in this summary
  • Male vs female ancestor proportions assume a single admixture pulse. Iron Gates dates rely on an assumed freshwater reservoir offset (545±70 yr), so continuous gene flow or a misestimated diet could bias sex-ratio and timing estimates. · noted in this summary

Read this if you study farmer-forager interactions in Neolithic Europe, Balkan or Pontic population history, or sex-biased admixture inferred by comparing X-chromosome and autosomal ancestry in ancient DNA.

Citations
708
FWCI
45.68
Field percentile
top 1%

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISLazaridis '16Hofmanová '16Feldman '19Lazaridis '14Mittnik '17Marchi '20
82% semantic matchGenetics
Iosif Lazaridis, Dani Nadel, Gary Rollefson+35 more·bioRxiv·Jun 16, 2016