1 POPULATION GENOMICS OF POSTGLACIAL WESTERN EURASIA 1 2 Morten E. Allentoft1,2*§, Martin Sikora1*§, Alba Refoyo-Martínez1§, Evan K. Irving-Pease1§, Anders Fischer1,3,4§, 3 William Barrie5§, Andrés Ingason6,1§, Jesper Stenderup1, Karl-Göran Sjögren3, Alice Pearson7, Bárbara Sousa da 4 Mota8,9, Bettina Schulz Paulsson3, Alma Halgren10, Ruairidh Macleod1,5,11, Marie Louise Schjellerup Jørkov12, Fabrice 5 Demeter1,13, Lasse Sørensen14, Poul Otto Nielsen14, Rasmus A. Henriksen1, Tharsika Vimala1, Hugh McColl1, Ashot 6 Margaryan15,16, Melissa Ilardo17, Andrew Vaughn18, Morten Fischer Mortensen14, Anne Birgitte Nielsen19, Mikkel 7 Ulfeldt Hede20, Niels Nørkjær Johannsen21, Peter Rasmussen14, Lasse Vinner1, Gabriel Renaud22, Aaron Stern18, Theis 8 Zetner Trolle Jensen15, Gabriele Scorrano1, Hannes Schroeder15, Per Lysdahl23, Abigail Daisy Ramsøe1, Andrei 9 Skorobogatov24, Andrew Joseph Schork6,25, Anders Rosengren6,1, Anthony Ruter1, Alan Outram26, Aleksey A. 10 Timoshenko27, Alexandra Buzhilova28, Alfredo Coppa29, Alisa Zubova30, Ana Maria Silva31,63, Anders J. Hansen1, 11 Andrey Gromov30, Andrey Logvin32, Anne Birgitte Gotfredsen1, Bjarne Henning Nielsen33, Borja González-Rabanal34, 12 Carles Lalueza-Fox35,36, Catriona J. McKenzie26, Charleen Gaunitz1, Concepción Blasco37, Corina Liesau37, Cristina 13 Martinez-Labarga38, Dmitri V. Pozdnyakov27, David Cuenca-Solana39,40, David O. Lordkipanidze41,42, Dmitri En’shin43, 14 Domingo C. Salazar-García44,45, T. Douglas Price3,46, Dušan Borić29,47, Elena Kostyleva48, Elizaveta V. Veselovskaya49, 15 Emma R. Usmanova50,51,52, Enrico Cappellini15, Erik Brinch Petersen53, Esben Kannegaard54, Francesca Radina55, Fulya 16 Eylem Yediay1, Henri Duday57, Igor Gutiérrez-Zugasti39, Ilya Merts58, Inna Potekhina59,60, Irina Shevnina32, Isin 17 Altinkaya1, Jean Guilaine61, Jesper Hansen62, Joan Emili Aura Tortosa44, João Zilhão63,64, Jorge Vega65, Kristoffer Buck 18 Pedersen66, Krzysztof Tunia67, Lei Zhao1, Liudmila N. Mylnikova27, Lars Larsson68, Laure Metz69, Levon 19 Yepiskoposyan70,99, Lisbeth Pedersen71, Lucia Sarti72, Ludovic Orlando73, Ludovic Slimak69, Lutz Klassen54, Malou 20 Blank3, Manuel González-Morales39, Mara Silvestrini74, Maria Vretemark75, Marina S. Nesterova27, Marina Rykun76, 21 Mario Federico Rolfo77, Marzena Szmyt78, Marcin Przybyła79, Mauro Calattini72, Mikhail Sablin80, Miluše 22 Dobisíková81, Morten Meldgaard82, Morten Johansen83, Natalia Berezina28, Nick Card84, Nikolai A. Saveliev85, Olga 23 Poshekhonova43, Olga Rickards38, Olga V. Lozovskaya86, Olivér Gábor87, Otto Christian Uldum83,88, Paola Aurino89, 24 Pavel Kosintsev90,91, Patrice Courtaud57, Patricia Ríos37, Peder Mortensen92†, Per Lotz93,94, Per Persson95, Pernille 25 Bangsgaard96, Peter de Barros Damgaard1, Peter Vang Petersen14, Pilar Prieto Martinez97, Piotr Włodarczak67, Roman 26 V. Smolyaninov98, Rikke Maring22,54, Roberto Menduiña65, Ruben Badalyan99, Rune Iversen53, Ruslan Turin24, Sergey 27 Vasilyev49,56, Sidsel Wåhlin23, Svetlana Borutskaya28, Svetlana Skochina43, Søren Anker Sørensen93, Søren H. 28 Andersen101, Thomas Jørgensen93, Yuri B. Serikov102, Vyacheslav I. Molodin27, Vaclav Smrcka103, Victor Merz104, 29 Vivek Appadurai6, Vyacheslav Moiseyev30, Yvonne Magnusson105, Kurt H. Kjær1, Niels Lynnerup12, Daniel J. 30 Lawson106, Peter H. Sudmant10,18, Simon Rasmussen107, Thorfinn Korneliussen1@, Richard Durbin7,108@, Rasmus 31 Nielsen10,1@, Olivier Delaneau8,9@, Thomas Werge1,6,109@, Fernando Racimo1@, Kristian Kristiansen1,3@, Eske 32 Willerslev1,5,110*@ 33 34 35 Affiliations 36 1Lundbeck Foundation GeoGenetics Centre, Globe Institute, University of Copenhagen, Copenhagen, Denmark. 2Trace 37 and Environmental DNA (TrEnD) Laboratory, School of Molecular and Life Sciences, Curtin University, Perth, 38 Australia. 3Department of Historical Studies, University of Gothenburg, Gothenburg, Sweden. 4Sealand Archaeology, 39 Gl. Roesnaesvej 27, 4400 Kalundborg, Denmark. 5GeoGenetics Group, Department of Zoology, University of 40 Cambridge, Cambridge, UK. 6Institute of Biological Psychiatry, Mental Health Services, Copenhagen University 41 Hospital, Roskilde, Denmark. 7Department of Genetics, University of Cambridge, Cambridge, UK. 8Department of 42 Computational Biology, University of Lausanne, Switzerland. 9Swiss Institute of Bioinformatics, University of 43 Lausanne, Switzerland. 10Department of Integrative Biology, University of California, Berkeley, USA. 11Research 44 department of Genetics, Evolution and Environment, University College London, London, UK. 12Laboratory of 45 Biological Anthropology, Department of Forensic Medicine, University of Copenhagen, Copenhagen, Denmark. 46 13Muséum national d’Histoire naturelle, CNRS, Université de Paris, Musée de l’Homme, Paris, France. 14The National 47 Museum of Denmark, Ny Vestergade 10, Copenhagen, Denmark. 15Section for Evolutionary Genomics, GLOBE 48 Institute, University of Copenhagen, Copenhagen, Denmark. 16Centre for Evolutionary Hologenomics, University of 49 Copenhagen, Copenhagen, Denmark. 17Anthropology Department, University of Utah, USA. 18Center for 50 Computational Biology, University of California, Berkeley, USA. 19Department of Geology, Lund University, Lund, 51 Sweden. 20Tårnby Gymnasium og HF, Kastrup, Denmark. 21Department of Archaeology and Heritage Studies, Aarhus 52 University, Aarhus, Denmark. 22Department of Health Technology, Section of Bioinformatics, Technical University of 53 Denmark, Kongens Lyngby, Denmark. 23Vendsyssel Historiske Museum, DK-10110 Hjørring, Denmark. 24Terra Ltd., 54 Letchik Zlobin St. 20, Voronezh, 397257, Russian Federation. 25Neurogenomics Division, The Translational Genomics 55 Research Institute (TGEN), Phoenix, AZ, USA. 26Department of Archaeology, University of Exeter, Exeter, UK. 56 27Institute of Archeology and Ethnography, Siberian Branch of the Russian Academy of Sciences, Novosibirsk, Russian 57 Federation. 28Department of Anthropology, Faculty of Biology, Lomonosov Moscow State University, Moscow, 58 Russian Federation. 29Department of Environmental Biology, Sapienza University of Rome, Rome, Italy. 30Peter the 59 2 Great Museum of Anthropology and Ethnography (Kunstkamera), Russian Academy of Sciences, Saint Petersburg, 60 Russian Federation. 31CIAS, Department of Life Science, University of Coimbra, Coimbra, Portugal. 32Kostanay 61 Regional University A. Baitursynov, Kostanay, Kazakhstan. 33Vesthimmerlands Museum, Søndergade 44, Aars, 62 Denmark. 34Grupo EvoAdapta, Departamento de Ciencias Históricas, Universidad de Cantabria, Santander, Spain. 63 35Institute of Evolutionary Biology, CSIC-Universitat Pompeu Fabra, Barcelona, Spain. 36Natural Sciences Museum of 64 Barcelona (MCNB), Barcelona, Spain. 37Departamento de Prehistoria y Arqueología, Universidad Autónoma de 65 Madrid, Madrid, Spain. 38Department of Biology, University of Rome "Tor Vergata", Rome, Italy. 39Instituto 66 Internacional de Investigaciones Prehistóricas de Cantabria, Universidad de Cantabria, Santander, Spain. 40Centre de 67 Recherche en Archéologie, Archeosciences, Histoire (CReAAH), UMR-6869 CNRS, Rennes, France. 41Georgian 68 National Museum, Tbilisi, Georgia. 42Tbilisi State University, Tbilisi, Georgia. 43IPND, Tyumen Scientific Centre, 69 Siberian Branch of the Russian Academy of Sciences, Tyumen, Russian Federation. 44Departament de Prehistòria, 70 Arqueologia i Història Antiga, Universitat de València, València, Spain. 45Department of Geological Sciences, 71 University of Cape Town, Cape Town, South Africa. 46Laboratory for Archaeological Chemistry, Department of 72 Anthropology, University of Wisconsin-Madison, Madison, USA. 47Department of Anthropology, New York 73 University, New York, USA. 48Institute of Humanities, Ivanovo State University, Ivanovo, Russian Federation. 74 49Institute of Ethnology and Anthropology, Russian Academy of Sciences, Moscow, Russian Federation. 50Saryarka 75 Archaeological Institute, Buketov Karaganda University, Karaganda, Kazakhstan. 51South Ural State University, 76 Chelyabinsk, Russia. 52A. Kh. Khalikov Institute of Archeology of the Academy of Sciences of the Republic of 77 Tatarstan. 53The Saxo Institute, University of Copenhagen, Copenhagen, Denmark. 54Museum Østjylland, 78 Stemannsgade 2, Randers, Denmark. 55Soprintendenza Archeologia Belle Arti e Paesaggio per la Città Metropolitana di 79 Bari, Via Pier l’Eremita, 25, 73122, Bari, Italy. 56Center for Egyptological studies, Russian Academy of Sciences, 80 Moscow, Russian Federation. 57UMR 51102 PACEA, CNRS, Université de Bordeaux, 33645 Pessac, France. 58A.Kh. 81 Margulan Institute of Archaeology, Almaty, Kazakhstan. 59Institute of Archaeology, National Academy of Sciences of 82 Ukraine, Kyiv, Ukraine. 60National University of Kyiv-Mohyla Academy, Kyiv, Ukraine. 61Collège de France, 78331 83 Paris cedex 05, France. 62Svendborg Museum, Grubbemøllevej 13, Svendborg, Denmark 63UNIARQ, University of 84 Lisbon, Lisbon, Portugal. 64ICREA, University of Barcelona, Barcelona, Spain. 65ARGEA Consultores SL, C. de San 85 Crispín, Madrid, Spain. 66Museum Sydøstdanmark, Algade 103, 4902 Vordingborg, Denmark. 67Institute of 86 Archaeology and Ethnology, Polish Academy of Sciences, Kraków, Poland. 68Department of Archaeology and Ancient 87 History, Lund University, Lund, Sweden. 69CNRS UMR 5938, Toulouse Jean Jaurès University, Maison de la 88 Recherche, 5 Allées Antonio Machado, 31091 Toulouse, Cedex 9, France. 70Institute of Molecular Biology, National 89 Academy of Sciences, Yerevan, Armenia. 71HistorieUdvikler, Gl. Roesnaesvej 27, DK-4400 Kalundborg, Denmark. 90 72Department of history and cultural heritage, University of Siena, Siena, Italy. 73Centre d'Anthropobiologie et de 91 Génomique de Toulouse, CNRS UMR 5500, Université Paul Sabatier, Toulouse, France. 74Soprintendenza per i Beni 92 Archeologici delle Marche, Via Birarelli 18, Ancona, Italy. 75Västergötlands museum, Stadsträdgården, Skara, Sweden. 93 76Cabinet of Anthropology, Tomsk State University, Tomsk, Russian Federation. 77Department of History, Humanities 94 and Society, University of Rome "Tor Vergata", Rome, Italy. 78Faculty of Archaeology, Adam Mickiewicz University 95 in Poznań, Poznań, Poland. 79Institute of Archaeology, Jagiellonian University, Kraków, Poland. 80Zoological Institute 96 of Russian Academy of Sciences, St. Petersburg, Russian Federation. 81Department of Anthropology, Czech National 97 Museum, Prague, Czech Republic. 82Department of Health and Nature, University of Greenland, Greenland. 83The 98 Viking Ship Museum, Vindeboder 12, Roskilde, Denmark. 84Archaeology Institute, University of Highlands and 99 Islands, Scotland, UK. 85Scientific Research Center “Baikal region”, Irkutsk State University; 1, K. Marx st., Irkutsk, 100 Russian Federation. 86Laboratory for Experimental Traceology, Institute for the History of Material Culture of the 101 Russian Academy of Sciences, St. Petersburg, Russian Federation. 87Janus Pannonius Museum, Pécs, Hungary. 102 88Langelands Museum, Jens Winthersvej 12, 5900 Rudkøbing, Denmark. 89Soprintendenza Archeologia, Belle Arti e 103 Paesaggio per la provincia di Cosenza, Cosenza, Italy. 90Paleoecology Laboratory, Institute of Plant and Animal 104 Ecology, Ural Branch of the Russian Academy of Sciences, Ekaterinburg, Russian Federation. 91Department of History 105 of the Institute of Humanities, Ural Federal University, Ekaterinburg, Russian Federation. 92Centre for the Study of 106 Early Agricultural Societies, Department of Cross-Cultural and Regional Studies, University of Copenhagen, 2300 107 Copenhagen, Denmark. 93Museum Nordsjælland, Frederiksgade 9, 3400 Hillerød. 94Museum Vestsjælland, 108 Klosterstræde 18, 4300 Holbæk, Denmark. 95Museum of Cultural History, University of Oslo, St. Olavs Plass NO-0130 109 Oslo, Norway. 96ArchaeoScience, GLOBE Institute, University of Copenhagen, Copenhagen, Denmark. 97Department 110 of History, University of Santiago de Compostela, Spain. 98Lipetsk Regional Scientific Public Organisation 111 "Archaeological Research", Lipetsk, Russian Federation. 99Institute of Archaeology and Ethnography, National 112 Academy of Sciences, Yerevan, Armenia. 100Russian-Armenian University, Yerevan, Armenia. 101Moesgaard Museum, 113 Moesgård Allé 15, Højbjerg, Denmark. 102Nizhny Tagil State Socio-Pedagogical Institute, Nizhny Tagil, Russia. 114 103Institute for History of Medicine, First Faculty of Medicine, Charles University, Prague, Czech Republic. 104Centre 115 for Archaeological Research Toraighyrov University, Pavlodar, Kazakhstan. 105Malmö Museer, Malmöhusvägen 6, 116 Malmö, Sweden. 106Institute of Statistical Sciences, School of Mathematics, University of Bristol, Bristol, UK. 107Novo 117 Nordisk Foundation Centre for Protein Research, Faculty of Health and Medical Sciences, University of Copenhagen, 118 Copenhagen N, Denmark. 108Wellcome Sanger Institute, Wellcome Genome Campus, Cambridge, UK. 109Department 119 3 of Clinical Medicine, University of Copenhagen, 2200 Copenhagen N, Denmark. 110MARUM Center for Marine 120 Environmental Sciences and Faculty of Geosciences, University of Bremen, Bremen, Germany. 121 122 * Corresponding authors; email: morten.allentoft@curtin.edu.au, martin.sikora@sund.ku.dk, ew493@cam.ac.uk 123 § These authors contributed equally to this work. 124 @ These authors jointly supervised this work. 125 126 † Deceased 8th December, 2022. 127 128 129 Summary 130 Western Eurasia witnessed several large-scale human migrations during the Holocene1–5. To 131 investigate the cross-continental impacts we shotgun-sequenced 317 primarily Mesolithic and 132 Neolithic genomes from across Northern and Western Eurasia. These were imputed alongside 133 published data to obtain diploid genotypes from >1,600 ancient humans. Our analyses 134 revealed a ‘Great Divide’ genomic boundary extending from the Black Sea to the Baltic. 135 Mesolithic hunter-gatherers (HGs) were highly genetically differentiated east and west of this 136 zone, and the impact of the neolithisation was equally disparate. Large-scale ancestry shifts 137 occurred in the west as farming was introduced, including near-total replacements of HGs in 138 many areas, whereas no substantial ancestry shifts happened east of the zone during the same 139 period. Similarly, relatedness decreased in the west from the Neolithic transition onwards, 140 while east of the Urals relatedness remained high until ~4,000 BP, consistent with persistence 141 of localised HG groups. The boundary dissolved when Yamnaya-related ancestry spread 142 across western Eurasia around 5,000 BP resulting in a second major turnover that reached 143 most parts of Europe within a 1,000-year span. The genetic origin and fate of the Yamnaya 144 have remained elusive but we demonstrate that HGs from the Middle Don region contributed 145 ancestry to them. Yamnaya-groups later admixed with individuals associated with the 146 Globular Amphora Culture before expanding into Europe. Similar turnovers occurred in 147 western Siberia, where we report new genomic data from a ‘Neolithic steppe’ cline spanning 148 the Siberian forest steppe to Lake Baikal. These prehistoric migrations had profound and 149 lasting effects on the genetic diversity of Eurasian populations. 150 151 Keywords: 152 Population genomics, ancient DNA, Mesolithic, Neolithic, Eurasia 153 154 Introduction 155 Genetic diversity in West Eurasian human populations was largely shaped by three major 156 prehistoric migrations: anatomically modern hunter-gatherers (HGs) occupying the area from c. 157 45,000 BP4,6; Neolithic farmers expanding from the Middle East from c. 11,000 BP4; and steppe 158 pastoralists coming out of the Pontic Steppe c. 5,000 BP1,2. Palaeogenomic analyses have 159 uncovered the early post-glacial colonisation routes7 resulting in a basal ancestral dichotomy 160 between HGs in central/western Europe and HG groups represented further east8. Western HG 161 (WHG) ancestry appears to be derived directly from ancestry sources related to Epigravettian, 162 Azilian and Epipalaeolithic cultures (the ‘Villabruna Cluster’)9, while Eastern HG (EHG) ancestry 163 shows additional admixture with an Upper Palaeolithic (UP) Siberian source (‘Ancient North 164 Eurasian’, ANE)10. The WHG ancestry composition was regionally variable in the Mesolithic 165 populations. There is evidence for continuous local admixture in Iberian HGs11, contrasting with a 166 more homogenous WHG ancestry profile in Britain and northwestern continental Europe, 167 suggesting ancestry formation prior to expansion12. The timing of the ancestry admixture that 168 formed EHG has been estimated at 13-15,000 BP and the composition seem to follow a cline 169 mailto:morten.allentoft@curtin.edu.au mailto:martin.sikora@sund.ku.dk mailto:ew482@cam.ac.uk https://paperpile.com/c/xoiDNc/WMvJ+ZmVr+bPGT+etCF+vReu https://paperpile.com/c/xoiDNc/etCF+zqE2 https://paperpile.com/c/xoiDNc/etCF https://paperpile.com/c/xoiDNc/WMvJ+ZmVr https://paperpile.com/c/xoiDNc/Gqy2p https://paperpile.com/c/xoiDNc/SIJpH https://paperpile.com/c/xoiDNc/ApbLg https://paperpile.com/c/xoiDNc/2U8A6 https://paperpile.com/c/xoiDNc/7SiAn https://paperpile.com/c/xoiDNc/7KgOt 4 broadly correlated with geography: Baltic and Ukrainian HGs showing more affinity to the 170 Villabruna UP cluster ancestry, compared to HGs in Russia who displayed more ANE5,7,13,14. 171 Genomic analyses of Mesolithic skeletal material from the Scandinavian Peninsula has revealed 172 varied mixes of WHG and EHG ancestry among the later Mesolithic populations3,15,16. 173 Beyond these broad scale characterizations our knowledge on Mesolithic population structure and 174 demographic admixture processes is limited and with significant chronological and geographic 175 information gaps. This is partly owing to a relative paucity of well-preserved Mesolithic human 176 skeletons older than 8,000 years and partly because most ancient DNA (aDNA) studies on the 177 Mesolithic and Neolithic periods have been restricted to individuals from Europe. The 178 archaeological record indicates a boundary from the eastern Baltic to the Black Sea, east of which 179 HG societies persisted for much longer than in western Europe despite similar distance to the 180 distribution centre for early agriculture in the Middle East17. Components of eastern and western 181 HG ancestry appear highly variable in this boundary region5,18,19 but the wider spatiotemporal 182 genetic implications of the east-west division is unclear. The spatiotemporal mapping of population 183 dynamics east of Europe, including Northern and Central Asia during the same time period is 184 limited. In these regions the term ‘Neolithic’ is characterised by cultural and economic changes 185 including societal network differences, changes in lithic technology and use of pottery. For 186 instance, the Neolithic cultures of the Central Asian steppe and the Russian taiga belt possessed 187 pottery, but retained a HG economy alongside stone blade technology similar to the preceding 188 Mesolithic cultures20. A fundamental lack of data from some key regions and periods has prohibited 189 a deeper understanding of how the neolithisation differed in timing, mechanisms, and impact across 190 Northern and Western Eurasia. 191 The transition from hunting and gathering to farming was based on domesticated plants and animals 192 of Middle Eastern origin, and represents one of the most fundamental shifts in demography, health, 193 lifestyle and culture in human prehistory. The neolithisation process in large parts of Europe was 194 accompanied by the arrival of immigrants of Anatolian descent21. For example, in Iberia the 195 Neolithic began with the abrupt spread of immigrant farmers of Anatolian-Aegean ancestry along 196 the Mediterranean and Atlantic coasts, after which admixture with local HGs gradually took place11. 197 Similarly, in Southeast and Central Europe farming rapidly spread with Anatolian Neolithic 198 farmers, who were to some extent subsequently admixed with local HGs22–27. Conversely, in Britain 199 data suggest a complete replacement of the HG population when agriculture was introduced by 200 incoming continental farmers and without a subsequent resurgence of local HG ancestry12,28. In the 201 East Baltic region a markedly different neolithisation trajectory occurred, with introduction of 202 domesticates only at the emergence of the Corded Ware Complex around 4,800 cal. BP18,19. 203 Similarly, in eastern Ukraine, HGs of Mesolithic ancestry co-existed for millennia with farming 204 groups further west5,29. These recent studies have all provided important regional contributions 205 towards understanding West Eurasian population history, but from a broader cross-continental 206 perspective, our knowledge is still patchy. A fundamental lack of data from some key regions and 207 periods has prohibited a deeper understanding of how neolithisation differed in timing, 208 mechanisms, and impact across Northern and Western Eurasia. 209 From approximately 5,000 BP, an ancestry component related to Early Bronze Age steppe 210 pastoralists such as the Yamnaya Culture rapidly spread across Europe through the expansion of the 211 Corded Ware Complex (CWC) and related cultures1,2. Although previous studies have identified 212 these large-scale migrations into Europe and Central Asia, central aspects concerning the 213 demographic processes are not resolved. Yamnaya ancestry (i.e. ‘steppe’ ancestry) has been 214 characterised broadly as a mix between EHG ancestry and Caucasus Hunter-Gatherer (CHG), 215 formed in a hypothetical admixture between a ‘Northern’ steppe source and a ‘Southern’ Caucasus 216 source30. However, the exact origins of these ancestry sources have not been identified. 217 Furthermore, with a few exceptions31–33 published Yamnaya Y-chromosomal haplogroups do not 218 match those found in post 5,000 BP Europeans and the origin of this patrilineal lineage is also 219 https://paperpile.com/c/xoiDNc/vReu+FDBJ3+aQBAm+Gqy2p https://paperpile.com/c/xoiDNc/bPGT+Tq94Z+0e4LP https://paperpile.com/c/xoiDNc/4i4mZ https://paperpile.com/c/xoiDNc/0jOeK+vReu+rcSxI https://paperpile.com/c/xoiDNc/JPwgX https://paperpile.com/c/xoiDNc/xasbG https://paperpile.com/c/xoiDNc/7SiAn https://paperpile.com/c/xoiDNc/MESKS+qcoGI+G7YOZ+kaBQi+VBGNr+enIVw https://paperpile.com/c/xoiDNc/kJqM4+7KgOt https://paperpile.com/c/xoiDNc/0jOeK+rcSxI https://paperpile.com/c/xoiDNc/vReu+yVuey https://paperpile.com/c/xoiDNc/WMvJ+ZmVr https://paperpile.com/c/xoiDNc/7YojO https://paperpile.com/c/xoiDNc/N4iBB+cdqn7+uiMLi 5 unresolved. Finally, in Europe ‘steppe’ ancestry has hitherto only been identified in admixed form, 220 but the origin of this admixture event and the mechanism by which the ancestry subsequently 221 spread with the CWC have remained elusive. 222 To investigate these formative processes at a cross-continental scale, we sequenced the genomes of 223 317 radiocarbon-dated (AMS) individuals of primarily Mesolithic and Neolithic origin, covering 224 major parts of Eurasia. We combined these with published shotgun-sequenced data to impute a 225 dataset of >1600 diploid ancient genomes. Of the 317 sampled ancient skeletons (Fig. 1, Extended 226 Data Fig. 1, Supplementary Data I) 272 were radiocarbon-dated within the project, while 30 dates 227 were derived from published literature, and 15 were dated by archaeological context. Dates were 228 corrected for marine and freshwater reservoir effects (Supplementary Note 4) and ranged from the 229 UP c. 25,700 calibrated years before present (cal. BP) to the mediaeval period (c. 1200 cal. BP). 230 However, 97% of the individuals (N=309) date to between 11,000 and 3,000 cal. BP, with a heavy 231 focus on individuals associated with various Mesolithic and Neolithic cultures. Geographically, the 232 317 sampled skeletons cover a vast territory across Eurasia, from Lake Baikal to the Atlantic coast 233 and from Scandinavia to the Middle East, deriving from contexts that include burial mounds, caves, 234 bogs and the seafloor (Supplementary Notes 6-7). Broadly, we can divide our research area into 235 three large regions: 1) central, western and northern Europe, 2) eastern Europe, including western 236 Russia, Belarus and Ukraine, and 3) the Urals and western Siberia (Supplementary Notes 6-7). 237 Samples cover many of the key Mesolithic and Neolithic cultures in Western Eurasia, such as the 238 Maglemose, Ertebølle, Funnel Beaker (TRB) and Corded Ware/Single Grave Cultures in 239 Scandinavia, the Cardial in the Mediterranean, the Körös and Linear Pottery (LBK) in SE and 240 Central Europe, and many archaeological cultures in the Ukraine, western Russia, and the trans-241 Ural (e.g. Veretye, Lyalovo, Volosovo, Kitoi). Our sampling was particularly dense in Denmark, 242 from where we present a detailed and continuous sequence of 100 genomes spanning the Early 243 Mesolithic to the Bronze Age (Allentoft, Sikora, Fischer et al. submitted*1). Dense sampling was 244 also obtained from Ukraine, Western Russia, and the trans-Ural, spanning from the Early 245 Mesolithic through the Neolithic, up to c. 5,000 BP. 246 247 248 Results and Discussion 249 Broad scale genetic structure 250 Ancient DNA was extracted from either dental cementum or petrous bones and the 317 genomes 251 were shotgun sequenced to a depth of coverage ranging between 0.01X and 7.1X (mean = 0.75X, 252 median = 0.26X), with >1X coverage for 81 genomes (Supplementary Note 1). We utilised a new 253 computational method optimised for low-coverage data34, to impute genotypes using the 1000 254 Genomes phased data35 as a reference panel. This was jointly applied to >1300 previously 255 published shotgun-sequenced genomes (Supplementary Data VII), resulting in a dataset of 8.5 256 million common SNPs (>1% Minor Allele Frequency (MAF) and imputation info score > 0.5) for 257 1,664 imputed diploid ancient genomes (Extended Data Fig. 2). For most downstream analyses 258 n=71 individuals were excluded as close relatives or with a contamination estimate >5%, resulting 259 in 1,593 genomes of which 1,492 were analysed as imputed (213 sequenced in this study) while 260 1* M.E.A., M. Sikora, A.F., K.-G.S., A.I., R. Macleod, A. Rosengren, B.S.P., M.L.S.J, Maria Novosolov, J.S., T.D.P., M.F.M., A.B.N., M.U.H., L.S., P.O.N., P.R., T.Z.T.J., A.R.-M., E.K.I.-P., W.B., W.B., A.P., B.S.d.M., F.D., R.A.H, T.V., H.M., A.V., L.V., G.R., A.J. Stern, N.N.J., A. Ramsøe, A.J. Schork, A. Ruter, A.B.G, B.H.N., E.B.P., E.K., J.H., K.B.P., L.P., L.K., M.M., M.J., O.C.U., P.L., P.B., P.V.P., R. Maring, R.I., S.W., S.A.S., S.H.A., T.J., N.L., D.J.L., S.R., T.S.K., K.H.K., R.D., F.R., R.N., O.D., T.W., K.K., E.W., 100 Ancient Genomes Show Two Rapid Population Turnovers in Neolithic Denmark. (submitted) https://paperpile.com/c/xoiDNc/4XCxT https://paperpile.com/c/xoiDNc/1NZld 6 n=101 were analysed as pseudo-haploids due to low coverage (<0.1X), and/or low imputation 261 quality (average genotype probability < 0.98). 262 263 We conducted a broad-scale characterization of this dataset using principal component analysis 264 (PCA) and model-based clustering (ADMIXTURE), which recapitulated previously described 265 ancestry clines in ancient Eurasian populations at increased resolution (Fig. 1; Extended Data Fig. 266 1; Supplementary Note 3d). Our imputed whole genomes allowed us to perform principal 267 component analysis using ancient genomes as input, instead of projecting onto a space defined by 268 modern variation. Strikingly, this resulted in much higher differentiation among the ancient 269 individuals than observed previously (Extended Data Fig. 1). This is particularly notable in a PCA 270 of West Eurasian individuals, where the variance explained by the first two PCs increases ~1.5 fold, 271 and present-day populations are confined within a small central area of the PCA space (Fig. 1d; 272 Extended Data Fig. 1c,d). These results are consistent with higher genetic differentiation between 273 ancient Europeans than present-day populations, reflecting more genetic isolation and lower 274 effective population sizes among ancient groups. 275 To obtain a finer-scale characterization of genetic ancestries across space and time, we employed an 276 approach akin to the widely used CHROMOPAINTER/FINESTRUCTURE workflow36–38. We first 277 performed community detection on a network constructed from pairwise identity-by-descent (IBD)-278 sharing similarities between ancient individuals to group them into hierarchically related clusters of 279 similar genetic ancestry (Extended Data Fig. 3; Supplementary Note 3c). At higher levels of the 280 hierarchy, the resulting clusters represented previously described ancestry groups reflecting broad 281 genetic structure, such as eastern and western European hunter-gatherers (“HG_EuropeE”, 282 “HG_EuropeW”; Extended Data Fig. 3). Clusters at the lowest level resolved fine-scale genetic 283 structure, grouping individuals within restricted spatiotemporal ranges and/or archaeological 284 contexts but also revealing previously unknown connections across broader geographical areas 285 (Extended Data Fig. 3; Supplementary Note 3f). These resulting clusters were subsequently used in 286 supervised ancestry modelling where sets of ‘target’ individuals were modelled as mixtures of 287 ‘source’ groups (Methods). 288 289 Post-LGM Population structure of HGs 290 Our study comprises 113 shotgun sequenced and imputed HG genomes of which 79 were 291 sequenced in this study. Among them, we report a 0.83X genome of an UP skeleton from Kotias 292 Klde Cave in Georgia, Caucasus (NEO283), directly dated to 26,052 - 25,323 cal. BP (95%). In the 293 PCA of all non-African individuals, it occupied a position distinct from other previously sequenced 294 UP individuals, shifted towards west Eurasians along PC1 (Supplementary Note 3d). Using 295 admixture graph modelling, we find that a well-fitting graph for this Caucasus UP lineage derives it 296 as a mixture of predominantly West Eurasian UP HG ancestry (76%) with ~24% contribution from 297 a ‘basal Eurasian’ ghost population, first observed in West Asian Neolithic individuals4 298 (Supplementary Note 3d, Fig. S3d.16). To further explore the fine-scale structure of later European 299 HGs, we then performed supervised ancestry modelling using sets of increasingly proximate source 300 clusters (Extended Data Fig. 4). We replicate previous results of broad-scale genetic differentiation 301 between HGs in eastern and western Europe after the Last Glacial Maximum (LGM)5,7. We show 302 that the deep ancestry divisions in the Eurasian human gene pool that were established during early 303 post-LGM dispersals7 persisted throughout the Mesolithic (Extended Data Fig. 4). Using distal sets 304 of pre-LGM HGs as sources, western HGs were modelled as predominantly derived from a source 305 related to the herein reported Caucasus UP individual from Kotias Klde cave (Caucasus_25000BP), 306 whereas eastern HGs showed varying amounts of ancestry related to a Siberian HG from Mal’ta 307 (Malta_24000BP; Extended Data Fig. 4a; Supplementary Data XII). Using post-LGM sources, this 308 divide is best represented by ancestry related to southern European (Italy_15000BP_9000 BP) and 309 https://paperpile.com/c/xoiDNc/7fCyi+C9mFL+1Q5YF https://paperpile.com/c/xoiDNc/etCF https://paperpile.com/c/xoiDNc/vReu+Gqy2p https://paperpile.com/c/xoiDNc/Gqy2p 7 Russian (RussiaNW_11000BP_8000BP) HGs, respectively, corresponding to the ‘WHG’ and 310 ‘EHG’ labels commonly used in previous studies. 311 Adding additional proximate sources allowed us to further refine the ancestry composition of 312 Northern European HGs. In Denmark, our 28 sequenced and imputed HG genomes derived almost 313 exclusively from a southern European source (Italy_15000BP_9000), with remarkable homogeneity 314 across a 5,000 year transect (Extended Data Fig. 4a; Supplementary Data XII) (Allentoft, Sikora, 315 Fischer et al. submitted). In contrast, we observed marked geographic variation in ancestry 316 composition of HGs from other parts of Scandinavia. Mesolithic individuals from Scandinavia were 317 broadly modelled as mixtures with varying proportions of eastern and western HGs using distal 318 post-LGM sources (“hgEur1”, Extended Data Fig. 4a), as previously reported15. In Mesolithic 319 individuals from southern Sweden, the eastern HG ancestry component was largely replaced by a 320 southeast European source (Romania_8800BP) in more proximate models, making up between 321 60%-70% of the ancestry (Extended Data Fig. 4a; Supplementary Data XII). Ancestry related to 322 Russian HGs increased in a cline towards the far north, peaking at ~75% in a late HG from Tromso 323 (VK531; ~4,350BP) (Extended Data Fig. 4a,c; Supplementary Data XII), also reflected in highest 324 IBD sharing of those individuals with Northern Russian HGs (Extended Data Fig. 4d). During the 325 late Mesolithic, we observed higher southern European HG ancestry in coastal individuals 326 (NEO260 from Evensås; NEO679 from Skateholm) than in earlier individuals from further inland. 327 Adding Danish HGs as proximate source substantially improved the fit for those two individuals 328 (“hgEur3”, Extended Data Fig. 4b), with estimated 58%-76% ancestry derived from Danish HGs 329 (“hgEur3”, Extended Data Fig.7a; Supplementary Data XII), suggesting a population genetic link 330 with Denmark where this ancestry prevailed (Extended Data Fig. 4c). These results indicate at least 331 three distinct waves of northwards HG ancestry into Scandinavia: (i) a predominantly southern 332 European source into Denmark and coastal southwestern Sweden; (ii) a source related to south-333 eastern European HGs into the Baltic and southeastern Sweden; and (iii) a northwest Russian 334 source into the far north, which then spread south along the Atlantic coast of Norway15 (Extended 335 Data Fig. 4c). These movements likely represent post glacial expansions from refugia areas shared 336 with many plant and animal species39. 337 On the Iberian Peninsula, the earliest individuals, including a ~9,200-year-old HG (NEO694) from 338 Santa Maira (eastern Spain), sequenced in this study, showed predominantly southern European HG 339 ancestry with a minor contribution from UP HG sources (Extended Data Fig. 4a). This observed UP 340 HG ancestry source mix likely reflects the pre-LGM Magdalenian-related ancestry component 341 previously reported in Iberian HGs11, for which a good source population proxy is lacking in our 342 dataset. In contrast, later individuals from Northern Iberia were more similar to HGs from 343 southeastern Europe, deriving ~30-40% of their ancestry from a source related to HGs from the 344 Balkans in more proximate models11,40 (Extended Data Fig. 4a; Supplementary Data XII). The 345 earliest evidence for this gene flow was observed in a Mesolithic individual from El Mazo, Spain 346 (NEO646) that was dated, calibrated and reservoir-corrected to c. 8,200 BP (8365-8182 cal. BP, 347 95%) but dated slightly earlier by context (8550-8330 BP41). The directly dated age coincides with 348 some of the oldest Mesolithic geometric microliths in northern Iberia, appearing around 8,200 BP at 349 this site41. An influx of southeastern European HG-related ancestry in Ukrainian individuals after 350 the Mesolithic (Extended Data Fig. 4a; Supplementary Data XII) suggests a similar eastwards 351 expansion in south-eastern Europe5. Interestingly, two newly reported ~7,300-year-old genomes 352 from the Middle Don River region in the Pontic-Caspian steppe (Golubaya Krinitsa, NEO113 & 353 NEO212) were found to be predominantly derived from earlier Ukrainian HGs, but with ~18-24% 354 of their ancestry contributed from a source related to HGs from the Caucasus 355 (Caucasus_13000BP_10000BP) (Extended Data Fig. 4a; Supplementary Data XII). Additional 356 lower coverage (non-imputed) genomes from the same site project in the same PCA space (Fig. 1d), 357 shifted away from the European HG cline towards Iran and the Caucasus. Using the linkage-358 disequilibrium-based method DATES42 we dated this admixture to ~8,300 BP (Supplementary Data 359 https://paperpile.com/c/xoiDNc/Tq94Z https://paperpile.com/c/xoiDNc/Tq94Z https://paperpile.com/c/xoiDNc/aV3e https://paperpile.com/c/xoiDNc/7SiAn https://paperpile.com/c/xoiDNc/IHJyU+7SiAn https://paperpile.com/c/xoiDNc/NwaUA https://paperpile.com/c/xoiDNc/NwaUA https://paperpile.com/c/xoiDNc/vReu https://paperpile.com/c/xoiDNc/k8W24 8 XIV). These results document genetic contact between populations from the Caucasus and the 360 steppe region much earlier than previously known, evidencing admixture prior to the advent of later 361 nomadic steppe cultures, in contrast to recent hypotheses, and further to the west than previously 362 reported5,43. 363 364 Major genetic transitions in Europe 365 Previous ancient genomics studies have documented multiple episodes of large-scale population 366 turnover in Europe within the last 10,000 yearse.g. 1,2,5,44 but the 317 novel genomes reported here fill 367 important knowledge gaps. Our analyses reveal profound differences in the spatiotemporal 368 neolithisation dynamics across Europe. Supervised admixture modelling (using the ‘deep’ ancestry 369 set; Supplementary Data IX) and spatiotemporal kriging45 document a broad east-west distinction 370 along a boundary zone running from the Black Sea to the Baltic. On the western side of this ‘Great 371 Divide’, the Neolithic transition is accompanied by large-scale shifts in genetic ancestry from local 372 HGs to farmers with Anatolian-related ancestry (Boncuklu_10000BP, Fig. 3a, Fig. 4; Extended 373 Data Figs. 5-7). The arrival of Anatolian-related ancestry in different regions spans an extensive 374 time period of over 3,000 years, from its earliest evidence in the Balkans (Lepenski Vir) at ~8,700 375 BP5 to c. 5,900 BP in Denmark. 376 Further, we corroborate previous reportse.g. 2,5,44,46 of widespread, low-level admixture between early 377 European farmers and local HGs resulting in a resurgence of HG ancestry in many regions of 378 Europe during subsequent centuries (Extended Data Fig. 8b,c; Supplementary Data XIII). The 379 resulting estimated HG ancestry proportions rarely exceeded 10%, with notable exceptions 380 observed in individuals from south-eastern Europe (Iron Gates), Sweden (Pitted Ware Culture) as 381 well as herein reported Early Neolithic genomes from Portugal (western Cardial), which are 382 estimated to harbour 27% – 43% Iberian HG ancestry (Iberia_9000BP_7000BP). The latter result, 383 together with an estimated admixture date of just ~200 years earlier (Supplementary Data XIV) 384 suggests extensive first-contact admixture, and is in agreement with archaeological inferences 385 derived from modelling the spread of farming along west Mediterranean Europe47. Neolithic 386 individuals from Denmark showed some of the highest overall HG ancestry proportions (up to 387 ~25%), but mostly derived from non-local Western European-related HGs 388 (EuropeW_13500BP_8000BP), with only a small contribution from local Danish HG groups in 389 some individuals (Extended Data Fig. 8b; Supplementary Note 3f). 390 We find evidence for regional stratification in early Neolithic farmer ancestries in subsequent 391 Neolithic groups. Specifically, southern European early farmers were found to have provided major 392 genetic ancestry to Neolithic groups of later dates in Western Europe, while central European early 393 farmer ancestry was mainly observed in subsequent Neolithic groups in eastern Europe and 394 Scandinavia (Extended Data Fig. 8e). These results are consistent with distinct migratory routes of 395 expanding farmer populations as previously suggested48. 396 On the eastern side of the ‘Great Divide’ no ancestry shifts can be observed during this period. In 397 the East Baltic region (see also49), the Ukraine and Western Russia, local HG ancestry prevailed 398 until ~5,000 BP without noticeable input of Anatolian-related farmer ancestry (Fig. 3-4; Extended 399 Data Figs. 5-7). This Eastern genetic continuity is in congruence with the archaeological record, 400 which shows persistence of pottery-using forager groups in this wide region, and a delayed 401 introduction of cultivation and animal husbandry by several thousand years (Supplementary Note 402 5). Around 5,000 BP major demographic events unfolded on the Eurasian Steppe resulting in 403 steppe-related ancestry spreading rapidly both eastwards and westwards1,2, marking the end of the 404 great population genomic divide (Fig. 4, Fig. 8). We find that this second transition happened at a 405 faster pace than during the neolithisation, reaching most parts of Europe within a ~1,000-year time 406 period after first appearing in the eastern Baltic region ~4,800 cal. BP (Fig. 3). In line with previous 407 reports we observe that by c. 4,200 cal. BP, steppe-related ancestry was already dominant in 408 https://paperpile.com/c/xoiDNc/vReu+oE5vc https://paperpile.com/c/xoiDNc/WMvJ+ZmVr+qS28Y+vReu/?prefix=e.g.,,,&noauthor=0,0,0,0 https://paperpile.com/c/xoiDNc/Qz5B6 https://paperpile.com/c/xoiDNc/vReu https://paperpile.com/c/xoiDNc/ZmVr+troZT+qS28Y+vReu/?prefix=e.g.,,, https://paperpile.com/c/xoiDNc/fJMYe https://paperpile.com/c/xoiDNc/KUnde https://paperpile.com/c/xoiDNc/wq3en https://paperpile.com/c/xoiDNc/WMvJ+ZmVr 9 individuals from Britain, France and the Iberian Peninsula12,50. Strikingly, because of the delayed 409 neolithisation in southern Scandinavia these dynamics resulted in two episodes of large-scale 410 genetic turnover in Denmark and southern Sweden within roughly a 1,000-year period (Fig. 3) 411 (Allentoft, Sikora, Fischer et al. submitted). 412 While the broader impacts of the steppe migrations around 5,000 cal. BP are well known, the origin 413 of this ancestry has remained a mystery. Here we demonstrate that the steppe ancestry composition 414 (Steppe_5000BP_4300BP) can be modelled as a mixture of ~65% ancestry related to herein 415 reported HG genomes from the Middle Don River region (MiddleDon_7500BP) and ~35% ancestry 416 related to HGs from Caucasus (Caucasus_13000BP_10000BP) (Extended Data Fig. 6; 417 Supplementary Data IX). Thus, Middle Don HGs, who already carried ancestry related to Caucasus 418 HGs (Extended Data Fig. 4a), serve as a hitherto unknown proximal source for the majority 419 ancestry contribution into Yamnaya-related genomes. The individuals in question derive from the 420 burial ground Golubaya Krinitsa (Supplementary Note 3). Material culture and burial practices at 421 this site are similar to the Mariupol-type graves, which are widely found in neighbouring regions of 422 the Ukraine, for instance along the Dnepr River. They belong to the group of complex pottery-using 423 HGs mentioned above, but the genetic composition at Golubaya Krinitsa is different from the 424 remaining Ukrainian sites (Fig 2A, Extended Data Fig. 5). Lazaridis et al.30 suggested a model for 425 the formation of Yamnaya ancestry that includes a ‘Northern’ steppe source (EHG+CHG ancestry) 426 and a ‘Southern’ Caucasus Chalcolithic source (CHG ancestry) but without identifying the exact 427 origin of these sources. The Middle Don genomes analysed here display the appropriate balance of 428 EHG/CHG ancestry, suggesting them as likely candidates for the missing Northern proximate 429 source for Yamnaya ancestry. 430 The dynamics of the continent-wide transition from Neolithic farmer ancestry to Steppe-related 431 ancestry also differs markedly between geographic regions. The contribution of local Neolithic 432 ancestry to the incoming groups was high in eastern, western and southern Europe, reaching >50% 433 on the Iberian Peninsula (“postNeol” set; Extended Data Fig. 6; Supplementary Data X)40. 434 Scandinavia, however, portrays a dramatically different picture, with much lower contributions 435 (<15%), including near-complete replacement of the local population in some regions (Extended 436 Data Fig. 9b). Steppe-related ancestry accompanies and spreads with the formation of the CWC 437 across Europe and our results provide new evidence on the foundational admixture event. 438 Individuals associated with the CWC carry a mix of steppe-related and Neolithic farmer-related 439 ancestry and we show that the latter can be modelled as deriving exclusively from a genetic cluster 440 associated with the Late Neolithic Globular Amphora Culture (GAC) (Poland_5000BP_4700BP), 441 and this ancestry co-occurred with steppe-related ancestry across all sampled European regions 442 (Fig. 5a; Extended Data Fig. 6). This suggests that the spread of steppe-related ancestry was 443 predominantly mediated through groups already admixed with GAC-related farmer groups of the 444 eastern European plains — an observation that has major implications for understanding the 445 emergence of the CWC. 446 A stylistic connection between GAC and CWC ceramics has long been suggested, including the use 447 of amphora-shaped vessels and the development of cord decoration patterns51. Moreover, shortly 448 before the emergence of the earliest CWC groups, eastern GAC and western Yamnaya groups 449 exchanged cultural elements in the forest-steppe transition zone northwest of the Black Sea, where 450 GAC ceramic amphorae and flint axes were included in Yamnaya burials, and the typical Yamnaya 451 use of ochre was included in GAC burials52, indicating close interaction between these groups. 452 Previous ancient genomic data from a few individuals suggested that this was limited to cultural 453 influences and not population admixture53. However, in the light of our new genetic evidence it 454 appears that this zone, and possibly other similar zones of contact between GAC and groups from 455 the steppe (e.g. Yamnaya), were key in the formation of the CWC through which steppe-related 456 ancestry and GAC-related ancestry co-dispersed far towards the west and the northcf. 54. This 457 resulted in regionally diverse situations of interaction and admixture14,32 but a significant part of the 458 https://paperpile.com/c/xoiDNc/7KgOt+eP3tk https://paperpile.com/c/xoiDNc/7YojO https://paperpile.com/c/xoiDNc/IHJyU https://paperpile.com/c/xoiDNc/ATIEq https://paperpile.com/c/xoiDNc/i0nfh https://paperpile.com/c/xoiDNc/gecKB https://paperpile.com/c/xoiDNc/KYSne/?prefix=cf. https://paperpile.com/c/xoiDNc/cdqn7+aQBAm 10 CWC dispersal happened through corridors of cultural and demic transmission which had been 459 established by the GAC during the preceding period33,55. Differences in Y-chromosomal 460 haplogroups between CWC and Yamnaya suggests that the currently published Yamnaya-461 associated genomes do not represent the most direct source for the steppe ancestry component in 462 CWC32,33. This notion was here supported by proximate ancestry modelling using published 463 genomes1 associated with Yamnaya or Afanasievo cultural contexts as separate sources, which 464 revealed a subtle increase in affinity for an Afanasievo-related source over a Yamnaya-related 465 source in early European steppe-ancestry carrying individuals before 3,000 cal. BP (Fig. 5b; 466 Extended Data Fig. 9d). The result confirms subtle population genomic structure in the population 467 associated with Yamnaya/Afanasievo, showing that more dense sampling across the steppe horizon 468 will be required to find the direct source(s) for steppe ancestry in early CWC. 469 470 HG resilience east of the Urals 471 In contrast to the significant number of ancient HG genomes from western Eurasia studied to date, 472 genomic data from HGs east of the Urals have remained sparse. As noted above, these regions are 473 characterised by an early introduction of pottery from areas further east and were inhabited by 474 complex forager societies with permanent and sometimes fortified settlements20,56. Here, we 475 substantially expand knowledge on ancient populations of this region by reporting new genomic 476 data from 38 individuals, 28 of which date to pottery-associated HG contexts between 8,300-5,000 477 cal. BP (Supplementary Data II). The majority of these genomes form a previously only sparsely 478 sampled13,42 ‘Neolithic steppe’ cline spanning the Siberian forest steppe zones of the Irtysh, Ishim, 479 Ob, and Yenisei River basins to the Lake Baikal region (Fig. 1c; Extended Data Fig. 1A, 3E). 480 Supervised admixture modelling (using the “deep” set of ancestry sources; Supplementary Data 481 IX) revealed contributions from three major sources in these HGs from east of the Urals: early West 482 Siberian HG ancestry (SteppeC_8300BP_7000BP) dominated in the western Forest Steppe; 483 Northeast Asian HG ancestry (Amur_7500BP) was highest at Lake Baikal; and Paleosiberian 484 ancestry (SiberiaNE_9800BP) was observed in a cline of decreasing proportions from northern 485 Lake Baikal westwards across the forest steppe13 (Extended Data Figs. 7, 10a). 486 We used these Neolithic HG clusters (“postNeol” ancestry source set, Extended Data Fig. 7) as 487 putative source groups in more proximal admixture modelling to investigate the spatiotemporal 488 dynamics of ancestry compositions across the steppe and the Lake Baikal region after the Neolithic 489 period. We replicate previously reported evidence for a genetic shift towards higher forest steppe 490 HG ancestry (source SteppeCE_7000BP_3600BP) in Late Neolithic and Early Bronze Age 491 individuals (LNBA) at Lake Baikal (clusters Baikal_5600BP_5400BP and 492 Baikal_4800BP_4200BP) 13,57. However, ancestry related to this cluster is also already observed at 493 ~7,000 BP in herein-reported Neolithic HG individuals both at Lake Baikal (NEO199, NEO200), 494 and along the Angara river to the north (NEO843) (Extended Data Fig. 7). Both male individuals at 495 Lake Baikal belonged to Y-chromosome haplogroup Q1b1, characteristic of the later LNBA groups 496 in the same region (Supplementary Note 3b, Figure S3b.5). Together with an early estimated 497 admixture time (~7,300 cal. BP upper bound) for the LNBA groups (Supplementary Data XIV), 498 these results suggest that gene flow between HGs of Lake Baikal and the south Siberian forest 499 steppe regions already occurred during the Eastern Early Neolithic, consistent with archaeological 500 interpretations of contact. In this region, bifacially flaked tools first appeared near Baikal58 from 501 where the technique spread far to the west. We find echoes of such bifacial flaking in 502 archaeological complexes (Shiderty 3, Borly, Sharbakty 1, Ust-Narym, etc.) in Northern and 503 Eastern Kazakhstan, around 6,500-6,000 cal. BP59,60. Here, Mesolithic cultural networks with 504 Southwest Asia have also been recorded, as evidenced by pebble and flint lithics known from 505 Southwest Asia cultures61. 506 https://paperpile.com/c/xoiDNc/0rRVt+uiMLi https://paperpile.com/c/xoiDNc/cdqn7+uiMLi https://paperpile.com/c/xoiDNc/WMvJ https://paperpile.com/c/xoiDNc/JPwgX+jdfpK https://paperpile.com/c/xoiDNc/FDBJ3+k8W24 https://paperpile.com/c/xoiDNc/FDBJ3 https://paperpile.com/c/xoiDNc/FDBJ3+H6wfP https://paperpile.com/c/xoiDNc/QSYBS https://paperpile.com/c/xoiDNc/UdWKv+fUutG https://paperpile.com/c/xoiDNc/xwg9F/?locator=108 11 Genomes reported here also shed light on the genetic origins of the Early Bronze Age Okunevo 507 Culture in the Minusinsk Basin in Southern Siberia. In contrast to previous results, we find no 508 evidence for Lake Baikal HG-related ancestry in the Okunevo13,57 when using our newly reported 509 Siberian forest steppe HG genomes jointly with Lake Baikal LNBA genomes as putative proximate 510 sources. Instead, we found that they originate from admixture of a forest steppe HG source (best 511 modelled as mixture of clusters Steppe_6700BP_4600BP and SteppeCE_7000BP_3600BP) and 512 steppe-related ancestry (Steppe_5300BP_4000BP; Extended Data Fig. 7, set “postBA”; 513 Supplementary Data XI). We date the admixture with steppe-related ancestry to ~4,600 BP 514 (Supplementary Data XIV), and found it to be modelled exclusively from an Afanasievo-related 515 source in proximate modelling separating the Yamnaya and Afanasievo steppe-ancestries (Extended 516 Data Figs. 9d, 10c,e). This is direct evidence for gene flow from peoples of the Afanasievo Culture 517 that were closely related to Yamnaya and existed near Altai and Minusinsk Basin during the era of 518 the steppe migrations1,57. 519 From around 3,700 cal. BP, individuals across the steppe and Lake Baikal regions display markedly 520 different ancestry profiles (Fig. 6; Extended Data Fig. 7, 9b). We document a sharp increase in non-521 local ancestries, with only limited ancestry contributions from local HGs. The early stages of this 522 transition are characterised by influx of steppe-related ancestry, which decays over time from its 523 peak of ~70% in the earliest individuals. Similar to the dynamics in western Eurasia, steppe-related 524 ancestry is here correlated with GAC-related farmer ancestry (Poland_5000BP_4700BP; Fig. 6; 525 Extended Data Fig. 10b), recapitulating previously documented gene flow from GAC groups into 526 steppe/forest steppe neighbouring groups and the eastward expansion of admixed Western steppe 527 pastoralists from the Sintashta and Andronovo complexes during the Bronze Age42,62. However, 528 GAC-related ancestry is notably absent in individuals of the Okunevo Culture, and individuals with 529 steppe ancestry after 3,700BP show slight excess in affinity to Yamnaya over Afanasievo in 530 proximate modelling (Extended Data Fig. 10d), providing further support for two distinct eastward 531 migrations of Western steppe pastoralists during the early (Yamnaya-related) and later (Sintashta, 532 Andronovo) Bronze Age. The later stages of the transition are characterised by increasing Central 533 Asian (Turkmenistan_7000 BP_5000BP) and Northeast Asian-related (Amur_7500BP) ancestry 534 components (Fig. 6; Extended Data Fig. 10b). Together, these results show that deeply structured 535 HG ancestry dominated the eastern Eurasian steppe substantially longer than in West Eurasia, 536 before successive waves of population expansions swept across the steppe within the last 4,000 537 years. These included a large-scale introduction of domesticated horse lineages concomitant with 538 new equestrian equipment and spoke-wheeled chariotry62,63, as well as the adoption of millet as 539 robust subsistence crop64. 540 Sociocultural insights 541 We used patterns of pairwise IBD sharing between individuals to examine our data for temporal 542 shifts in relatedness within genetic clusters. We found clear trends of a reduction of within-cluster 543 relatedness over time, in both western and eastern Eurasia (Extended Data Fig. 11a). This pattern is 544 consistent with a scenario of increasing effective population sizes during this period65. 545 Nevertheless, we observe notable differences in temporal relatedness patterns between western and 546 eastern Eurasia, mirroring the wider difference in population dynamics discussed above. In the 547 west, within-group relatedness changed substantially during the Neolithic transition (~9,000 to 548 ~6,000 BP), where clusters of individuals with Anatolian farmer-related ancestry show overall 549 reduced IBD sharing compared to clusters of individuals with HG-associated ancestry (Extended 550 Data Fig. 11a). In the east, genetic relatedness remained high until ~4,000 BP, consistent with a 551 much longer persistence of smaller localised HG groups (Fig. 6; Extended Data Fig. 11a). 552 Next, we examined the data for evidence of recent parental relatedness, by identifying individuals 553 harbouring > 50cM of their genomes in long (>20cM) ROH segments66. We only detect 29 such 554 https://paperpile.com/c/xoiDNc/H6wfP+FDBJ3 https://paperpile.com/c/xoiDNc/WMvJ+H6wfP https://paperpile.com/c/xoiDNc/BKLr5+k8W24 https://paperpile.com/c/xoiDNc/BKLr5+zdZ0A https://paperpile.com/c/xoiDNc/KB48D https://paperpile.com/c/xoiDNc/bGEQl https://paperpile.com/c/xoiDNc/EZqg6 12 individuals out of a total sample of 1,396 imputed ancient genomes from across Eurasia (Extended 555 Data Fig. 11b). This suggests that close kin mating was not common in the regions and periods 556 covered by our data. No obviously discernible spatiotemporal or cultural clustering was observed 557 among the individuals with recent parental relatedness. Interestingly, a ~1,700-year-old Sarmatian 558 individual from Temyaysovo (tem003)67 was found homozygous for almost the entirety of 559 chromosome 2, but without evidence of ROHs elsewhere in the genome, suggesting the first 560 documented case of uniparental disomy in an ancient individual (Extended Data Fig. 11c). Among 561 several noteworthy familial relationships (see Supplementary Fig. S3c.2), we report a Mesolithic 562 father/son burial at Ertebølle (NEO568/NEO569), as well as a Mesolithic mother/daughter burial at 563 Dragsholm (NEO732/NEO733), Denmark (see also Allentoft, Sikora, Fischer et al. submitted). 564 565 Formation and dissolution of the divide 566 We have demonstrated the existence of a clear east-west genetic division extending from the Black 567 Sea to the Baltic, mirroring archaeological observations, and persisting over several millennia. We 568 show that this deep ancestry division in the Eurasian human gene pool that was established during 569 early post-LGM dispersals7 was maintained throughout the Mesolithic and Neolithic ages (Fig. 8). 570 Accordingly, we show that the genetic impact of the Neolithic transition was highly distinct east 571 and west of this boundary. These observations raise a series of questions related to understanding 572 the underlying drivers. 573 In eastern Europe, the expansion of Neolithic farming was halted for around 3,000 years and 574 this delay may be linked to environmental factors, with regions east of the division having more 575 continental climates and harsher winters, possibly less suited for Middle Eastern agricultural 576 practices68. Here, highly developed HG societies persisted with stable, complex and sometimes 577 fortified settlements, long distance exchange and large cemeteries69,70. A diet including freshwater 578 fish is clear both from our isotopic data (Supplementary Data II) and from analyses of lipids in 579 pottery70. In the northern forested regions of this boundary zone, HG societies persisted until the 580 emergence of the CWC around 5,000 cal. BP, whereas in the southern and eastern steppe regions, 581 hunting and gathering was eventually complemented with some animal husbandry (cattle and 582 sheep), and possibly horse herding in central Asia71. Some of these groups, such as Khvalynsk at 583 the Volga saw the emergence of male sodalities involved in wide-ranging trade connections of 584 copper objects from east central Europe and the Caucasus29. Settlements were confined mainly to 585 the flat floodplains and river valleys, while the steppe belt remained largely unexploited. 586 The eventual dissolution of this genetic, economic, and social border was driven by events 587 unfolding in the steppe region. Here, two temporal phases of technological innovations can be 588 observed archaeologically: the widespread dispersal of ox-drawn wheeled vehicles around 5,500 589 cal. BP and the later development of horse riding, and possible changing environmental 590 conditions72. This opened up the steppe as an economic zone, allowing Yamnaya groups to exploit 591 the steppe as pastoral nomads around 5,000 cal. BP73and all Eneolithic settlements along river 592 valleys were replaced by this new mobile economy74 finally dissolved the great genomic boundary 593 that had persisted in the preceding millenia (Fig. 8). 594 By 4,000 cal. BP the invention of chariot warfare and the adoption of millet as a food crop allowed 595 the final eastward expansion into central Asia and beyond by Andronovo and related groups, with 596 global legacies for the expansion of Indo-European languages75. Our study has provided new 597 genetic knowledge on these steppe migrations on two levels: we have identified a hitherto unknown 598 source of ancestry in HGs from the Middle Don region contributing ancestry to the steppe 599 pastoralists, and we have documented how the later spread of steppe-related ancestry into Europe 600 via the CWC was first mediated through peoples associated with the GAC. In a contact zone that 601 included forested northern regions, the CWC was rapidly formed from a cultural and genetic 602 amalgamation of steppe-groups related to Yamnaya and the GAC groups in eastern Europe. In 603 https://paperpile.com/c/xoiDNc/JCbLN https://paperpile.com/c/xoiDNc/Gqy2p https://paperpile.com/c/xoiDNc/9VYSZ https://paperpile.com/c/xoiDNc/5VhwE+MWpPo https://paperpile.com/c/xoiDNc/MWpPo https://paperpile.com/c/xoiDNc/IeoLi https://paperpile.com/c/xoiDNc/yVuey https://paperpile.com/c/xoiDNc/aNblH https://paperpile.com/c/xoiDNc/AMW9u https://paperpile.com/c/xoiDNc/yuFgi https://paperpile.com/c/xoiDNc/sgQ30 13 accordance with their mixed cultural and genetic background, the CWC practised a mixed 604 economy, employing various subsistence strategies in different environments. This flexibility would 605 have contributed substantially to their success in settling and adapting to very different ecological 606 and climatic settings over a very short period of time33. 607 608 References 609 1. Allentoft, M. E. et al. Population genomics of Bronze Age Eurasia. Nature 522, 167–172 (2015). 610 2. Haak, W. et al. Massive migration from the steppe was a source for Indo-European languages in Europe. Nature 611 522, 207–211 (2015). 612 3. Lazaridis, I. et al. Ancient human genomes suggest three ancestral populations for present-day Europeans. Nature 613 513, 409–413 (2014). 614 4. Lazaridis, I. et al. Genomic Insights into the Origin of Farming in the Ancient Near East. Nature 536, 419–24 615 (2016). 616 5. Mathieson, I. et al. The genomic history of southeastern Europe. Nature 555, 197–203 (2018). 617 6. Posth, C. et al. Pleistocene Mitochondrial Genomes Suggest a Single Major Dispersal of Non-Africans and a Late 618 Glacial Population Turnover in Europe. Curr. Biol. 26, 827–833 (2016). 619 7. Posth, C. et al. Palaeogenomics of Upper Palaeolithic to Neolithic European hunter-gatherers. Nature 615, 117–620 126 (2023). 621 8. Mathieson, I. et al. Genome-wide patterns of selection in 230 ancient Eurasians. Nature 528, 499–503 (2015). 622 9. Fu, Q. et al. The genetic history of Ice Age Europe. Nature 534, 200–205 (2016). 623 10. Raghavan, M. et al. Upper Palaeolithic Siberian genome reveals dual ancestry of Native Americans. Nature 505, 624 87–91 (2014). 625 11. Villalba-Mouco, V. et al. Survival of Late Pleistocene Hunter-Gatherer Ancestry in the Iberian Peninsula. Curr. 626 Biol. 29, 1169–1177.e7 (2019). 627 12. Brace, S. et al. Ancient genomes indicate population replacement in Early Neolithic Britain. Nat Ecol Evol 3, 765–628 771 (2019). 629 13. de Barros Damgaard, P. et al. The first horse herders and the impact of early Bronze Age steppe expansions into 630 Asia. Science 360, (2018). 631 14. Saag, L. et al. Genetic ancestry changes in Stone to Bronze Age transition in the East European plain. Sci Adv 7, 632 (2021). 633 15. Günther, T. et al. Population genomics of Mesolithic Scandinavia: Investigating early postglacial migration routes 634 and high-latitude adaptation. PLoS Biol. 16, e2003703 (2018). 635 16. Kashuba, N. et al. Ancient DNA from mastics solidifies connection between material culture and genetics of 636 mesolithic hunter–gatherers in Scandinavia. Communications Biology 2, 1–10 (2019). 637 17. Zvelebil, M., Domanska, L. & Dennell, R. Harvesting the Sea, Farming the Forest: The Emergence of Neolithic 638 Societies in the Baltic Region. (Bloomsbury Publishing, 1998). 639 18. Jones, E. R. et al. The Neolithic Transition in the Baltic Was Not Driven by Admixture with Early European 640 Farmers. Curr. Biol. 27, 576–582 (2017). 641 19. Mittnik, A. et al. The Genetic Prehistory of the Baltic Sea Region. Nat. Commun. 9, 442 (2018). 642 20. Kislenko, A. & Tatarintseva, N. The eastern Ural steppe at the end of the Stone Age. Late prehistoric exploitation 643 https://paperpile.com/c/xoiDNc/uiMLi http://paperpile.com/b/xoiDNc/WMvJ http://paperpile.com/b/xoiDNc/WMvJ http://paperpile.com/b/xoiDNc/WMvJ http://paperpile.com/b/xoiDNc/WMvJ http://paperpile.com/b/xoiDNc/WMvJ http://paperpile.com/b/xoiDNc/WMvJ http://paperpile.com/b/xoiDNc/WMvJ http://paperpile.com/b/xoiDNc/ZmVr http://paperpile.com/b/xoiDNc/ZmVr http://paperpile.com/b/xoiDNc/ZmVr http://paperpile.com/b/xoiDNc/ZmVr http://paperpile.com/b/xoiDNc/ZmVr http://paperpile.com/b/xoiDNc/ZmVr http://paperpile.com/b/xoiDNc/ZmVr http://paperpile.com/b/xoiDNc/ZmVr http://paperpile.com/b/xoiDNc/bPGT http://paperpile.com/b/xoiDNc/bPGT http://paperpile.com/b/xoiDNc/bPGT http://paperpile.com/b/xoiDNc/bPGT http://paperpile.com/b/xoiDNc/bPGT http://paperpile.com/b/xoiDNc/bPGT http://paperpile.com/b/xoiDNc/bPGT http://paperpile.com/b/xoiDNc/bPGT http://paperpile.com/b/xoiDNc/etCF http://paperpile.com/b/xoiDNc/etCF http://paperpile.com/b/xoiDNc/etCF http://paperpile.com/b/xoiDNc/etCF http://paperpile.com/b/xoiDNc/etCF http://paperpile.com/b/xoiDNc/etCF http://paperpile.com/b/xoiDNc/etCF http://paperpile.com/b/xoiDNc/etCF http://paperpile.com/b/xoiDNc/vReu http://paperpile.com/b/xoiDNc/vReu http://paperpile.com/b/xoiDNc/vReu http://paperpile.com/b/xoiDNc/vReu http://paperpile.com/b/xoiDNc/vReu http://paperpile.com/b/xoiDNc/vReu http://paperpile.com/b/xoiDNc/vReu http://paperpile.com/b/xoiDNc/zqE2 http://paperpile.com/b/xoiDNc/zqE2 http://paperpile.com/b/xoiDNc/zqE2 http://paperpile.com/b/xoiDNc/zqE2 http://paperpile.com/b/xoiDNc/zqE2 http://paperpile.com/b/xoiDNc/zqE2 http://paperpile.com/b/xoiDNc/zqE2 http://paperpile.com/b/xoiDNc/zqE2 http://paperpile.com/b/xoiDNc/Gqy2p http://paperpile.com/b/xoiDNc/Gqy2p http://paperpile.com/b/xoiDNc/Gqy2p http://paperpile.com/b/xoiDNc/Gqy2p http://paperpile.com/b/xoiDNc/Gqy2p http://paperpile.com/b/xoiDNc/Gqy2p http://paperpile.com/b/xoiDNc/Gqy2p http://paperpile.com/b/xoiDNc/Gqy2p http://paperpile.com/b/xoiDNc/SIJpH http://paperpile.com/b/xoiDNc/SIJpH http://paperpile.com/b/xoiDNc/SIJpH http://paperpile.com/b/xoiDNc/SIJpH http://paperpile.com/b/xoiDNc/SIJpH http://paperpile.com/b/xoiDNc/SIJpH http://paperpile.com/b/xoiDNc/SIJpH http://paperpile.com/b/xoiDNc/ApbLg http://paperpile.com/b/xoiDNc/ApbLg http://paperpile.com/b/xoiDNc/ApbLg http://paperpile.com/b/xoiDNc/ApbLg http://paperpile.com/b/xoiDNc/ApbLg http://paperpile.com/b/xoiDNc/ApbLg http://paperpile.com/b/xoiDNc/ApbLg http://paperpile.com/b/xoiDNc/2U8A6 http://paperpile.com/b/xoiDNc/2U8A6 http://paperpile.com/b/xoiDNc/2U8A6 http://paperpile.com/b/xoiDNc/2U8A6 http://paperpile.com/b/xoiDNc/2U8A6 http://paperpile.com/b/xoiDNc/2U8A6 http://paperpile.com/b/xoiDNc/2U8A6 http://paperpile.com/b/xoiDNc/2U8A6 http://paperpile.com/b/xoiDNc/7SiAn http://paperpile.com/b/xoiDNc/7SiAn http://paperpile.com/b/xoiDNc/7SiAn http://paperpile.com/b/xoiDNc/7SiAn http://paperpile.com/b/xoiDNc/7SiAn http://paperpile.com/b/xoiDNc/7SiAn http://paperpile.com/b/xoiDNc/7SiAn http://paperpile.com/b/xoiDNc/7SiAn http://paperpile.com/b/xoiDNc/7KgOt http://paperpile.com/b/xoiDNc/7KgOt http://paperpile.com/b/xoiDNc/7KgOt http://paperpile.com/b/xoiDNc/7KgOt http://paperpile.com/b/xoiDNc/7KgOt http://paperpile.com/b/xoiDNc/7KgOt http://paperpile.com/b/xoiDNc/7KgOt http://paperpile.com/b/xoiDNc/7KgOt http://paperpile.com/b/xoiDNc/FDBJ3 http://paperpile.com/b/xoiDNc/FDBJ3 http://paperpile.com/b/xoiDNc/FDBJ3 http://paperpile.com/b/xoiDNc/FDBJ3 http://paperpile.com/b/xoiDNc/FDBJ3 http://paperpile.com/b/xoiDNc/FDBJ3 http://paperpile.com/b/xoiDNc/FDBJ3 http://paperpile.com/b/xoiDNc/FDBJ3 http://paperpile.com/b/xoiDNc/aQBAm http://paperpile.com/b/xoiDNc/aQBAm http://paperpile.com/b/xoiDNc/aQBAm http://paperpile.com/b/xoiDNc/aQBAm http://paperpile.com/b/xoiDNc/aQBAm http://paperpile.com/b/xoiDNc/aQBAm http://paperpile.com/b/xoiDNc/aQBAm http://paperpile.com/b/xoiDNc/aQBAm http://paperpile.com/b/xoiDNc/Tq94Z http://paperpile.com/b/xoiDNc/Tq94Z http://paperpile.com/b/xoiDNc/Tq94Z http://paperpile.com/b/xoiDNc/Tq94Z http://paperpile.com/b/xoiDNc/Tq94Z http://paperpile.com/b/xoiDNc/Tq94Z http://paperpile.com/b/xoiDNc/Tq94Z http://paperpile.com/b/xoiDNc/Tq94Z http://paperpile.com/b/xoiDNc/0e4LP http://paperpile.com/b/xoiDNc/0e4LP http://paperpile.com/b/xoiDNc/0e4LP http://paperpile.com/b/xoiDNc/0e4LP http://paperpile.com/b/xoiDNc/0e4LP http://paperpile.com/b/xoiDNc/0e4LP http://paperpile.com/b/xoiDNc/0e4LP http://paperpile.com/b/xoiDNc/0e4LP http://paperpile.com/b/xoiDNc/4i4mZ http://paperpile.com/b/xoiDNc/4i4mZ http://paperpile.com/b/xoiDNc/4i4mZ http://paperpile.com/b/xoiDNc/4i4mZ http://paperpile.com/b/xoiDNc/0jOeK http://paperpile.com/b/xoiDNc/0jOeK http://paperpile.com/b/xoiDNc/0jOeK http://paperpile.com/b/xoiDNc/0jOeK http://paperpile.com/b/xoiDNc/0jOeK http://paperpile.com/b/xoiDNc/0jOeK http://paperpile.com/b/xoiDNc/0jOeK http://paperpile.com/b/xoiDNc/0jOeK http://paperpile.com/b/xoiDNc/rcSxI http://paperpile.com/b/xoiDNc/rcSxI http://paperpile.com/b/xoiDNc/rcSxI http://paperpile.com/b/xoiDNc/rcSxI http://paperpile.com/b/xoiDNc/rcSxI http://paperpile.com/b/xoiDNc/rcSxI http://paperpile.com/b/xoiDNc/rcSxI http://paperpile.com/b/xoiDNc/JPwgX http://paperpile.com/b/xoiDNc/JPwgX 14 of the Eurasian steppe 183–216 (1999). 644 21. Furholt, M. Mobility and Social Change: Understanding the European Neolithic Period after the Archaeogenetic 645 Revolution. Journal of Archaeological Research 29, 481–535 (2021). 646 22. Lipson, M. et al. Ancient genomes document multiple waves of migration in Southeast Asian prehistory. Science 647 361, 92–95 (2018). 648 23. Fernandes, D. M. et al. A genomic Neolithic time transect of hunter-farmer admixture in central Poland. Sci. Rep. 649 8, 14879 (2018). 650 24. Immel, A., Rinne, C., Meadows, J., Barquera, R. & Szolek, A. Neolithic genomes reveal a distinct ancient hla 651 allele pool and population transformation in Europe. bioRxiv (2019). 652 25. Jeong, C. et al. The genetic history of admixture across inner Eurasia. Nat Ecol Evol 3, 966–976 (2019). 653 26. Nikitin, A. G. et al. Interactions between earliest Linearbandkeramik farmers and central European hunter 654 gatherers at the dawn of European Neolithization. Sci. Rep. 9, 19544 (2019). 655 27. Gelabert, P. et al. Social and genetic diversity among the first farmers of Central Europe. bioRxiv 656 2023.07.07.548126 (2023) doi:10.1101/2023.07.07.548126. 657 28. Cassidy, L. M. et al. Neolithic and Bronze Age migration to Ireland and establishment of the insular Atlantic 658 genome. Proc. Natl. Acad. Sci. U. S. A. 113, 368–373 (2016). 659 29. Penske, S. et al. Early contact between late farming and pastoralist societies in southeastern Europe. Nature (2023) 660 doi:10.1038/s41586-023-06334-8. 661 30. Lazaridis, I. et al. The genetic history of the Southern Arc: A bridge between West Asia and Europe. Science 377, 662 eabm4247 (2022). 663 31. Egfjord, A. F.-H. et al. Genomic Steppe ancestry in skeletons from the Neolithic Single Grave Culture in 664 Denmark. PLoS One 16, e0244872 (2021). 665 32. Papac, L. et al. Dynamic changes in genomic and social structures in third millennium BCE central Europe. Sci 666 Adv 7, (2021). 667 33. Heyd, V. The mobility and migration revolution in 3rd millennium BC Europe. in Rethinking Migrations in Late 668 Prehistoric Eurasia 41–62 (Oxford University Press, 2023). 669 34. Rubinacci, S., Ribeiro, D. M., Hofmeister, R. J. & Delaneau, O. Efficient phasing and imputation of low-coverage 670 sequencing data using large reference panels. Nat. Genet. 53, 412 (2021). 671 35. 1000 Genomes Project Consortium et al. A global reference for human genetic variation. Nature 526, 68–74 672 (2015). 673 36. Leslie, S. et al. The fine-scale genetic structure of the British population. Nature 519, 309–314 (2015). 674 37. Hofmanová, Z. et al. Early farmers from across Europe directly descended from Neolithic Aegeans. Proc. Natl. 675 Acad. Sci. U. S. A. 113, 6886–6891 (2016). 676 38. Busby, G. B. et al. Admixture into and within sub-Saharan Africa. Elife 5, (2016). 677 39. Schmitt, T. Molecular biogeography of Europe: Pleistocene cycles and postglacial trends. Front. Zool. 4, 11 678 (2007). 679 40. Olalde, I. et al. The genomic history of the Iberian Peninsula over the past 8000 years. Science 363, 1230–1234 680 (2019). 681 41. García-Escárzaga, A. et al. Human forager response to abrupt climate change at 8.2 ka on the Atlantic coast of 682 Europe. Sci. Rep. 12, 6481 (2022). 683 http://paperpile.com/b/xoiDNc/JPwgX http://paperpile.com/b/xoiDNc/JPwgX http://paperpile.com/b/xoiDNc/xasbG http://paperpile.com/b/xoiDNc/xasbG http://paperpile.com/b/xoiDNc/xasbG http://paperpile.com/b/xoiDNc/xasbG http://paperpile.com/b/xoiDNc/xasbG http://paperpile.com/b/xoiDNc/xasbG http://paperpile.com/b/xoiDNc/MESKS http://paperpile.com/b/xoiDNc/MESKS http://paperpile.com/b/xoiDNc/MESKS http://paperpile.com/b/xoiDNc/MESKS http://paperpile.com/b/xoiDNc/MESKS http://paperpile.com/b/xoiDNc/MESKS http://paperpile.com/b/xoiDNc/MESKS http://paperpile.com/b/xoiDNc/MESKS http://paperpile.com/b/xoiDNc/qcoGI http://paperpile.com/b/xoiDNc/qcoGI http://paperpile.com/b/xoiDNc/qcoGI http://paperpile.com/b/xoiDNc/qcoGI http://paperpile.com/b/xoiDNc/qcoGI http://paperpile.com/b/xoiDNc/qcoGI http://paperpile.com/b/xoiDNc/qcoGI http://paperpile.com/b/xoiDNc/qcoGI http://paperpile.com/b/xoiDNc/G7YOZ http://paperpile.com/b/xoiDNc/G7YOZ http://paperpile.com/b/xoiDNc/G7YOZ http://paperpile.com/b/xoiDNc/G7YOZ http://paperpile.com/b/xoiDNc/kaBQi http://paperpile.com/b/xoiDNc/kaBQi http://paperpile.com/b/xoiDNc/kaBQi http://paperpile.com/b/xoiDNc/kaBQi http://paperpile.com/b/xoiDNc/kaBQi http://paperpile.com/b/xoiDNc/kaBQi http://paperpile.com/b/xoiDNc/kaBQi http://paperpile.com/b/xoiDNc/VBGNr http://paperpile.com/b/xoiDNc/VBGNr http://paperpile.com/b/xoiDNc/VBGNr http://paperpile.com/b/xoiDNc/VBGNr http://paperpile.com/b/xoiDNc/VBGNr http://paperpile.com/b/xoiDNc/VBGNr http://paperpile.com/b/xoiDNc/VBGNr http://paperpile.com/b/xoiDNc/VBGNr http://paperpile.com/b/xoiDNc/enIVw http://paperpile.com/b/xoiDNc/enIVw http://paperpile.com/b/xoiDNc/enIVw http://paperpile.com/b/xoiDNc/enIVw http://paperpile.com/b/xoiDNc/enIVw http://paperpile.com/b/xoiDNc/enIVw http://dx.doi.org/10.1101/2023.07.07.548126 http://dx.doi.org/10.1101/2023.07.07.548126 http://paperpile.com/b/xoiDNc/kJqM4 http://paperpile.com/b/xoiDNc/kJqM4 http://paperpile.com/b/xoiDNc/kJqM4 http://paperpile.com/b/xoiDNc/kJqM4 http://paperpile.com/b/xoiDNc/kJqM4 http://paperpile.com/b/xoiDNc/kJqM4 http://paperpile.com/b/xoiDNc/kJqM4 http://paperpile.com/b/xoiDNc/kJqM4 http://paperpile.com/b/xoiDNc/yVuey http://paperpile.com/b/xoiDNc/yVuey http://paperpile.com/b/xoiDNc/yVuey http://paperpile.com/b/xoiDNc/yVuey http://paperpile.com/b/xoiDNc/yVuey http://paperpile.com/b/xoiDNc/yVuey http://dx.doi.org/10.1038/s41586-023-06334-8 http://dx.doi.org/10.1038/s41586-023-06334-8 http://paperpile.com/b/xoiDNc/7YojO http://paperpile.com/b/xoiDNc/7YojO http://paperpile.com/b/xoiDNc/7YojO http://paperpile.com/b/xoiDNc/7YojO http://paperpile.com/b/xoiDNc/7YojO http://paperpile.com/b/xoiDNc/7YojO http://paperpile.com/b/xoiDNc/7YojO http://paperpile.com/b/xoiDNc/7YojO http://paperpile.com/b/xoiDNc/N4iBB http://paperpile.com/b/xoiDNc/N4iBB http://paperpile.com/b/xoiDNc/N4iBB http://paperpile.com/b/xoiDNc/N4iBB http://paperpile.com/b/xoiDNc/N4iBB http://paperpile.com/b/xoiDNc/N4iBB http://paperpile.com/b/xoiDNc/N4iBB http://paperpile.com/b/xoiDNc/N4iBB http://paperpile.com/b/xoiDNc/cdqn7 http://paperpile.com/b/xoiDNc/cdqn7 http://paperpile.com/b/xoiDNc/cdqn7 http://paperpile.com/b/xoiDNc/cdqn7 http://paperpile.com/b/xoiDNc/cdqn7 http://paperpile.com/b/xoiDNc/cdqn7 http://paperpile.com/b/xoiDNc/cdqn7 http://paperpile.com/b/xoiDNc/cdqn7 http://paperpile.com/b/xoiDNc/uiMLi http://paperpile.com/b/xoiDNc/uiMLi http://paperpile.com/b/xoiDNc/uiMLi http://paperpile.com/b/xoiDNc/uiMLi http://paperpile.com/b/xoiDNc/4XCxT http://paperpile.com/b/xoiDNc/4XCxT http://paperpile.com/b/xoiDNc/4XCxT http://paperpile.com/b/xoiDNc/4XCxT http://paperpile.com/b/xoiDNc/4XCxT http://paperpile.com/b/xoiDNc/4XCxT http://paperpile.com/b/xoiDNc/1NZld http://paperpile.com/b/xoiDNc/1NZld http://paperpile.com/b/xoiDNc/1NZld http://paperpile.com/b/xoiDNc/1NZld http://paperpile.com/b/xoiDNc/1NZld http://paperpile.com/b/xoiDNc/1NZld http://paperpile.com/b/xoiDNc/1NZld http://paperpile.com/b/xoiDNc/1NZld http://paperpile.com/b/xoiDNc/7fCyi http://paperpile.com/b/xoiDNc/7fCyi http://paperpile.com/b/xoiDNc/7fCyi http://paperpile.com/b/xoiDNc/7fCyi http://paperpile.com/b/xoiDNc/7fCyi http://paperpile.com/b/xoiDNc/7fCyi http://paperpile.com/b/xoiDNc/7fCyi http://paperpile.com/b/xoiDNc/C9mFL http://paperpile.com/b/xoiDNc/C9mFL http://paperpile.com/b/xoiDNc/C9mFL http://paperpile.com/b/xoiDNc/C9mFL http://paperpile.com/b/xoiDNc/C9mFL http://paperpile.com/b/xoiDNc/C9mFL http://paperpile.com/b/xoiDNc/C9mFL http://paperpile.com/b/xoiDNc/C9mFL http://paperpile.com/b/xoiDNc/1Q5YF http://paperpile.com/b/xoiDNc/1Q5YF http://paperpile.com/b/xoiDNc/1Q5YF http://paperpile.com/b/xoiDNc/1Q5YF http://paperpile.com/b/xoiDNc/1Q5YF http://paperpile.com/b/xoiDNc/1Q5YF http://paperpile.com/b/xoiDNc/1Q5YF http://paperpile.com/b/xoiDNc/aV3e http://paperpile.com/b/xoiDNc/aV3e http://paperpile.com/b/xoiDNc/aV3e http://paperpile.com/b/xoiDNc/aV3e http://paperpile.com/b/xoiDNc/aV3e http://paperpile.com/b/xoiDNc/aV3e http://paperpile.com/b/xoiDNc/IHJyU http://paperpile.com/b/xoiDNc/IHJyU http://paperpile.com/b/xoiDNc/IHJyU http://paperpile.com/b/xoiDNc/IHJyU http://paperpile.com/b/xoiDNc/IHJyU http://paperpile.com/b/xoiDNc/IHJyU http://paperpile.com/b/xoiDNc/IHJyU http://paperpile.com/b/xoiDNc/IHJyU http://paperpile.com/b/xoiDNc/NwaUA http://paperpile.com/b/xoiDNc/NwaUA http://paperpile.com/b/xoiDNc/NwaUA http://paperpile.com/b/xoiDNc/NwaUA http://paperpile.com/b/xoiDNc/NwaUA http://paperpile.com/b/xoiDNc/NwaUA http://paperpile.com/b/xoiDNc/NwaUA http://paperpile.com/b/xoiDNc/NwaUA 15 42. Narasimhan, V. M. et al. The formation of human populations in South and Central Asia. Science 365, (2019). 684 43. Wang, C.-C. et al. Ancient human genome-wide data from a 3000-year interval in the Caucasus corresponds with 685 eco-geographic regions. Nat. Commun. 10, 590 (2019). 686 44. Lipson, M. et al. Parallel palaeogenomic transects reveal complex genetic history of early European farmers. 687 Nature 551, 368 (2017). 688 45. Racimo, F. et al. The Spatiotemporal Spread of Human Migrations during the European Holocene. Proc. Natl. 689 Acad. Sci. U. S. A. 117, 8989–9000 (2020). 690 46. Martiniano, R. et al. The population genomics of archaeological transition in west Iberia: Investigation of ancient 691 substructure using imputation and haplotype-based methods. PLoS Genet. 13, e1006852 (2017). 692 47. Isern, N., Zilhão, J., Fort, J. & Ammerman, A. J. Modeling the role of voyaging in the coastal spread of the Early 693 Neolithic in the West Mediterranean. Proc. Natl. Acad. Sci. U. S. A. 114, 897–902 (2017). 694 48. Betti, L. et al. Climate shaped how Neolithic farmers and European hunter-gatherers interacted after a major 695 slowdown from 6,100 BCE to 4,500 BCE. Nat Hum Behav 4, 1004–1010 (2020). 696 49. Saag, L. et al. Extensive Farming in Estonia Started through a Sex-Biased Migration from the Steppe. Curr. Biol. 697 27, 2185–2193.e6 (2017). 698 50. Seguin-Orlando, A. et al. Heterogeneous Hunter-Gatherer and Steppe-Related Ancestries in Late Neolithic and 699 Bell Beaker Genomes from Present-Day France. Curr. Biol. 31, 1072–1083.e10 (2021). 700 51. Furholt, M. Die Złota-Gruppe in Kleinpolen: Ein Beispiel für die Transformation eines Zeichensystems? 701 Germania: Anzeiger der Römisch-Germanischen Kommission des Deutschen Archäologischen Instituts 86, 1–28 702 (2008). 703 52. Szmyt, M. In the Reaches Far of Two Worlds: On the Study of Contacts between the Societies of the Globular 704 Amphora and Yamnaya Cultures. in A Turning of Ages: Jubilee Book Dedicated to Professor Jan Machnik on His 705 70th Anniversary (ed. Kadrow, S.) 443–466 (Institute of Archaeology and Ethnology, Polish Academy of 706 Sciences, Cracow Branch, 2000). 707 53. Tassi, F. et al. Genome diversity in the Neolithic Globular Amphorae culture and the spread of Indo-European 708 languages. Proc. Biol. Sci. 284, (2017). 709 54. Nordqvist, K. & Heyd, V. The Forgotten Child of the Wider Corded Ware Family: Russian Fatyanovo Culture in 710 Context. Proceedings of the Prehistoric Society 86, 65–93 (2020). 711 55. Kristiansen, K. et al. Re-theorising mobility and the formation of culture and language among the Corded Ware 712 Culture in Europe. Antiquity 91, 334–347 (2017). 713 56. Borzunov, V. A. The neolithic fortified settlements of the Western Siberia and Trans-Urals. Rossijskaâ arheologiâ 714 20–34 (2013). 715 57. Yu, H. et al. Paleolithic to Bronze Age Siberians Reveal Connections with First Americans and across Eurasia. 716 Cell 181, 1232–1245.e20 (2020). 717 58. Okladnikov, A. P. Neolit i bronzovyi vek Pribaikaliya [Neolithic and Bronze Age of the Baikal region]. 411 (AS 718 USSR Publications, 1950). 719 59. Merz, V. On the origin of the complexes of the Baikal type in the Eneolithic of Kazakhstan. in Paleodemography 720 and migration processes in Western Siberia in antiquity and the Middle Ages (ed. Kiryushin, Y. F.) 39–42 (Altai 721 State University, 1994). 722 60. Merz, V. Periodization of the Holocene complexes of Northern and Central Kazakhstan based on the materials of 723 http://paperpile.com/b/xoiDNc/k8W24 http://paperpile.com/b/xoiDNc/k8W24 http://paperpile.com/b/xoiDNc/k8W24 http://paperpile.com/b/xoiDNc/k8W24 http://paperpile.com/b/xoiDNc/k8W24 http://paperpile.com/b/xoiDNc/k8W24 http://paperpile.com/b/xoiDNc/k8W24 http://paperpile.com/b/xoiDNc/oE5vc http://paperpile.com/b/xoiDNc/oE5vc http://paperpile.com/b/xoiDNc/oE5vc http://paperpile.com/b/xoiDNc/oE5vc http://paperpile.com/b/xoiDNc/oE5vc http://paperpile.com/b/xoiDNc/oE5vc http://paperpile.com/b/xoiDNc/oE5vc http://paperpile.com/b/xoiDNc/oE5vc http://paperpile.com/b/xoiDNc/qS28Y http://paperpile.com/b/xoiDNc/qS28Y http://paperpile.com/b/xoiDNc/qS28Y http://paperpile.com/b/xoiDNc/qS28Y http://paperpile.com/b/xoiDNc/qS28Y http://paperpile.com/b/xoiDNc/qS28Y http://paperpile.com/b/xoiDNc/qS28Y http://paperpile.com/b/xoiDNc/qS28Y http://paperpile.com/b/xoiDNc/Qz5B6 http://paperpile.com/b/xoiDNc/Qz5B6 http://paperpile.com/b/xoiDNc/Qz5B6 http://paperpile.com/b/xoiDNc/Qz5B6 http://paperpile.com/b/xoiDNc/Qz5B6 http://paperpile.com/b/xoiDNc/Qz5B6 http://paperpile.com/b/xoiDNc/Qz5B6 http://paperpile.com/b/xoiDNc/Qz5B6 http://paperpile.com/b/xoiDNc/troZT http://paperpile.com/b/xoiDNc/troZT http://paperpile.com/b/xoiDNc/troZT http://paperpile.com/b/xoiDNc/troZT http://paperpile.com/b/xoiDNc/troZT http://paperpile.com/b/xoiDNc/troZT http://paperpile.com/b/xoiDNc/troZT http://paperpile.com/b/xoiDNc/troZT http://paperpile.com/b/xoiDNc/fJMYe http://paperpile.com/b/xoiDNc/fJMYe http://paperpile.com/b/xoiDNc/fJMYe http://paperpile.com/b/xoiDNc/fJMYe http://paperpile.com/b/xoiDNc/fJMYe http://paperpile.com/b/xoiDNc/fJMYe http://paperpile.com/b/xoiDNc/KUnde http://paperpile.com/b/xoiDNc/KUnde http://paperpile.com/b/xoiDNc/KUnde http://paperpile.com/b/xoiDNc/KUnde http://paperpile.com/b/xoiDNc/KUnde http://paperpile.com/b/xoiDNc/KUnde http://paperpile.com/b/xoiDNc/KUnde http://paperpile.com/b/xoiDNc/KUnde http://paperpile.com/b/xoiDNc/wq3en http://paperpile.com/b/xoiDNc/wq3en http://paperpile.com/b/xoiDNc/wq3en http://paperpile.com/b/xoiDNc/wq3en http://paperpile.com/b/xoiDNc/wq3en http://paperpile.com/b/xoiDNc/wq3en http://paperpile.com/b/xoiDNc/wq3en http://paperpile.com/b/xoiDNc/wq3en http://paperpile.com/b/xoiDNc/eP3tk http://paperpile.com/b/xoiDNc/eP3tk http://paperpile.com/b/xoiDNc/eP3tk http://paperpile.com/b/xoiDNc/eP3tk http://paperpile.com/b/xoiDNc/eP3tk http://paperpile.com/b/xoiDNc/eP3tk http://paperpile.com/b/xoiDNc/eP3tk http://paperpile.com/b/xoiDNc/eP3tk http://paperpile.com/b/xoiDNc/ATIEq http://paperpile.com/b/xoiDNc/ATIEq http://paperpile.com/b/xoiDNc/ATIEq http://paperpile.com/b/xoiDNc/ATIEq http://paperpile.com/b/xoiDNc/ATIEq http://paperpile.com/b/xoiDNc/ATIEq http://paperpile.com/b/xoiDNc/ATIEq http://paperpile.com/b/xoiDNc/i0nfh http://paperpile.com/b/xoiDNc/i0nfh http://paperpile.com/b/xoiDNc/i0nfh http://paperpile.com/b/xoiDNc/i0nfh http://paperpile.com/b/xoiDNc/i0nfh http://paperpile.com/b/xoiDNc/i0nfh http://paperpile.com/b/xoiDNc/gecKB http://paperpile.com/b/xoiDNc/gecKB http://paperpile.com/b/xoiDNc/gecKB http://paperpile.com/b/xoiDNc/gecKB http://paperpile.com/b/xoiDNc/gecKB http://paperpile.com/b/xoiDNc/gecKB http://paperpile.com/b/xoiDNc/gecKB http://paperpile.com/b/xoiDNc/gecKB http://paperpile.com/b/xoiDNc/KYSne http://paperpile.com/b/xoiDNc/KYSne http://paperpile.com/b/xoiDNc/KYSne http://paperpile.com/b/xoiDNc/KYSne http://paperpile.com/b/xoiDNc/KYSne http://paperpile.com/b/xoiDNc/KYSne http://paperpile.com/b/xoiDNc/0rRVt http://paperpile.com/b/xoiDNc/0rRVt http://paperpile.com/b/xoiDNc/0rRVt http://paperpile.com/b/xoiDNc/0rRVt http://paperpile.com/b/xoiDNc/0rRVt http://paperpile.com/b/xoiDNc/0rRVt http://paperpile.com/b/xoiDNc/0rRVt http://paperpile.com/b/xoiDNc/0rRVt http://paperpile.com/b/xoiDNc/jdfpK http://paperpile.com/b/xoiDNc/jdfpK http://paperpile.com/b/xoiDNc/jdfpK http://paperpile.com/b/xoiDNc/jdfpK http://paperpile.com/b/xoiDNc/H6wfP http://paperpile.com/b/xoiDNc/H6wfP http://paperpile.com/b/xoiDNc/H6wfP http://paperpile.com/b/xoiDNc/H6wfP http://paperpile.com/b/xoiDNc/H6wfP http://paperpile.com/b/xoiDNc/H6wfP http://paperpile.com/b/xoiDNc/H6wfP http://paperpile.com/b/xoiDNc/H6wfP http://paperpile.com/b/xoiDNc/QSYBS http://paperpile.com/b/xoiDNc/QSYBS http://paperpile.com/b/xoiDNc/QSYBS http://paperpile.com/b/xoiDNc/QSYBS http://paperpile.com/b/xoiDNc/UdWKv http://paperpile.com/b/xoiDNc/UdWKv http://paperpile.com/b/xoiDNc/UdWKv http://paperpile.com/b/xoiDNc/UdWKv http://paperpile.com/b/xoiDNc/UdWKv http://paperpile.com/b/xoiDNc/fUutG 16 the multilayer site Shiderty 3. (Kemerovo State University, 2008). 724 61. Merts, V. K. Neolithization Processes in the Northeast Kazakhstan. Herald of Omsk University. Series ‘Historical 725 Studies’ 3, 99–109 (2018). 726 62. Damgaard, P. de B. et al. 137 ancient human genomes from across the Eurasian steppes. Nature 557, 369–374 727 (2018). 728 63. Librado, P. et al. The origins and spread of domestic horses from the Western Eurasian steppes. Nature 598, 634–729 640 (2021). 730 64. Huang, Y. et al. The early adoption of East Asian crops in West Asia: rice and broomcorn millet in northern Iran. 731 Antiquity 1–16 (2023). 732 65. Palamara, P. F., Lencz, T., Darvasi, A. & Pe’er, I. Length distributions of identity by descent reveal fine-scale 733 demographic history. Am. J. Hum. Genet. 91, 809–822 (2012). 734 66. Ringbauer, H., Novembre, J. & Steinrücken, M. Parental relatedness through time revealed by runs of 735 homozygosity in ancient DNA. Nat. Commun. 12, 5425 (2021). 736 67. Krzewińska, M. et al. Ancient genomes suggest the eastern Pontic-Caspian steppe as the source of western Iron 737 Age nomads. Sci Adv 4, eaat4457 (2018). 738 68. Matuzeviciute, G. M. The Possible Geographic Margin Effect on the Delay of Agriculture Introduction in the East 739 Baltic. Eesti Arheoloogia Ajakiri 22, 149–162 (2018). 740 69. Piezonka, H. Jäger, Fischer, Töpfer: Wildbeutergruppen mit früher Keramik in Nordosteuropa im 6. und 5. 741 Jahrtausend v. Chr. (Archäologie in Eurasien). (Habelt, 2015). 742 70. Oras, E. et al. The adoption of pottery by north-east European hunter-gatherers: Evidence from lipid residue 743 analysis. J. Archaeol. Sci. 78, 112–119 (2017). 744 71. Matuzeviciute, G. M. et al. Archaeobotanical investigations at the earliest horse herder site of Botai in Kazakhstan. 745 Archaeol. Anthropol. Sci. 11, 6243–6258 (2019). 746 72. Anthony, D. W. Pontic-Caspian Mesolithic and Early Neolithic societies at the time of the Black Sea flood: a small 747 audience and small effects. The Black Sea flood question: changes in (2007). 748 73. Trautmann, M. et al. First bioanthropological evidence for Yamnaya horsemanship. Sci Adv 9, eade2451 (2023). 749 74. Anthony, D. W. et al. The Eneolithic cemetery at Khvalynsk on the Volga River. Praehistorische Zeitschrift 97, 750 22–67 (2022). 751 75. Kristiansen, K., Kroonen, G. & Willerslev, E. The Indo-European Puzzle Revisited: Integrating Archaeology, 752 Genetics, and Linguistics. (Cambridge University Press, 2023). 753 754 755 Figure Legends 756 Fig. 1: Sample overview and broad scale genetic structure. (A), (B) Geographic and temporal distribution of the 317 757 ancient genomes sequenced and reported in this study. Insert shows dense sampling in Denmark (see Allentoft, Sikora, 758 Fischer et al., submitted). Age and geographic region of ancient individuals are indicated by plot symbol colour and 759 shape, respectively. Colour scale for age is capped at 15,000 years, older individuals are indicated with black colour. 760 Random jitter was added to geographic coordinates to avoid overplotting. (C), (D) Principal component analysis of 761 3,316 modern and ancient individuals from Eurasia, Oceania, and the Americas (C), and restricted to 2,126 individuals 762 from western Eurasia (west of the Urals) (D). Principal components were defined using both modern and imputed 763 ancient (n=1492) genomes passing all filters, with the remaining low-coverage ancient genomes projected. Ancient 764 genomes sequenced in this study are indicated with black circles (imputed genomes passing all filters, n=213) or grey 765 http://paperpile.com/b/xoiDNc/fUutG http://paperpile.com/b/xoiDNc/xwg9F http://paperpile.com/b/xoiDNc/xwg9F http://paperpile.com/b/xoiDNc/xwg9F http://paperpile.com/b/xoiDNc/xwg9F http://paperpile.com/b/xoiDNc/xwg9F http://paperpile.com/b/xoiDNc/xwg9F http://paperpile.com/b/xoiDNc/BKLr5 http://paperpile.com/b/xoiDNc/BKLr5 http://paperpile.com/b/xoiDNc/BKLr5 http://paperpile.com/b/xoiDNc/BKLr5 http://paperpile.com/b/xoiDNc/BKLr5 http://paperpile.com/b/xoiDNc/BKLr5 http://paperpile.com/b/xoiDNc/BKLr5 http://paperpile.com/b/xoiDNc/BKLr5 http://paperpile.com/b/xoiDNc/zdZ0A http://paperpile.com/b/xoiDNc/zdZ0A http://paperpile.com/b/xoiDNc/zdZ0A http://paperpile.com/b/xoiDNc/zdZ0A http://paperpile.com/b/xoiDNc/zdZ0A http://paperpile.com/b/xoiDNc/zdZ0A http://paperpile.com/b/xoiDNc/zdZ0A http://paperpile.com/b/xoiDNc/zdZ0A http://paperpile.com/b/xoiDNc/KB48D http://paperpile.com/b/xoiDNc/KB48D http://paperpile.com/b/xoiDNc/KB48D http://paperpile.com/b/xoiDNc/KB48D http://paperpile.com/b/xoiDNc/KB48D http://paperpile.com/b/xoiDNc/KB48D http://paperpile.com/b/xoiDNc/bGEQl http://paperpile.com/b/xoiDNc/bGEQl http://paperpile.com/b/xoiDNc/bGEQl http://paperpile.com/b/xoiDNc/bGEQl http://paperpile.com/b/xoiDNc/bGEQl http://paperpile.com/b/xoiDNc/bGEQl http://paperpile.com/b/xoiDNc/EZqg6 http://paperpile.com/b/xoiDNc/EZqg6 http://paperpile.com/b/xoiDNc/EZqg6 http://paperpile.com/b/xoiDNc/EZqg6 http://paperpile.com/b/xoiDNc/EZqg6 http://paperpile.com/b/xoiDNc/EZqg6 http://paperpile.com/b/xoiDNc/JCbLN http://paperpile.com/b/xoiDNc/JCbLN http://paperpile.com/b/xoiDNc/JCbLN http://paperpile.com/b/xoiDNc/JCbLN http://paperpile.com/b/xoiDNc/JCbLN http://paperpile.com/b/xoiDNc/JCbLN http://paperpile.com/b/xoiDNc/JCbLN http://paperpile.com/b/xoiDNc/JCbLN http://paperpile.com/b/xoiDNc/9VYSZ http://paperpile.com/b/xoiDNc/9VYSZ http://paperpile.com/b/xoiDNc/9VYSZ http://paperpile.com/b/xoiDNc/9VYSZ http://paperpile.com/b/xoiDNc/9VYSZ http://paperpile.com/b/xoiDNc/9VYSZ http://paperpile.com/b/xoiDNc/5VhwE http://paperpile.com/b/xoiDNc/5VhwE http://paperpile.com/b/xoiDNc/5VhwE http://paperpile.com/b/xoiDNc/5VhwE http://paperpile.com/b/xoiDNc/MWpPo http://paperpile.com/b/xoiDNc/MWpPo http://paperpile.com/b/xoiDNc/MWpPo http://paperpile.com/b/xoiDNc/MWpPo http://paperpile.com/b/xoiDNc/MWpPo http://paperpile.com/b/xoiDNc/MWpPo http://paperpile.com/b/xoiDNc/MWpPo http://paperpile.com/b/xoiDNc/MWpPo http://paperpile.com/b/xoiDNc/IeoLi http://paperpile.com/b/xoiDNc/IeoLi http://paperpile.com/b/xoiDNc/IeoLi http://paperpile.com/b/xoiDNc/IeoLi http://paperpile.com/b/xoiDNc/IeoLi http://paperpile.com/b/xoiDNc/IeoLi http://paperpile.com/b/xoiDNc/IeoLi http://paperpile.com/b/xoiDNc/IeoLi http://paperpile.com/b/xoiDNc/aNblH http://paperpile.com/b/xoiDNc/aNblH http://paperpile.com/b/xoiDNc/aNblH http://paperpile.com/b/xoiDNc/aNblH http://paperpile.com/b/xoiDNc/AMW9u http://paperpile.com/b/xoiDNc/AMW9u http://paperpile.com/b/xoiDNc/AMW9u http://paperpile.com/b/xoiDNc/AMW9u http://paperpile.com/b/xoiDNc/AMW9u http://paperpile.com/b/xoiDNc/AMW9u http://paperpile.com/b/xoiDNc/AMW9u http://paperpile.com/b/xoiDNc/yuFgi http://paperpile.com/b/xoiDNc/yuFgi http://paperpile.com/b/xoiDNc/yuFgi http://paperpile.com/b/xoiDNc/yuFgi http://paperpile.com/b/xoiDNc/yuFgi http://paperpile.com/b/xoiDNc/yuFgi http://paperpile.com/b/xoiDNc/yuFgi http://paperpile.com/b/xoiDNc/yuFgi http://paperpile.com/b/xoiDNc/sgQ30 http://paperpile.com/b/xoiDNc/sgQ30 http://paperpile.com/b/xoiDNc/sgQ30 http://paperpile.com/b/xoiDNc/sgQ30 17 diamonds (pseudo-haploid projected genomes, n=104). Genomes of modern individuals are shown in grey, with 766 population labels corresponding to their median coordinates. 767 768 Fig. 2: Genetic ancestry transects of Western Eurasia. (a) Regional timelines of genetic ancestry compositions 769 within the past 12,000 years in western Eurasia. Ancestry proportions in 1,012 imputed ancient genomes (representing 770 populations west of the Urals) inferred using supervised ancestry modelling with the “deep” HG ancestry source 771 groups. Coloured bars within the timelines represent ancestry proportions for temporally consecutive individuals, with 772 the width corresponding to their age difference. Individuals with identical age were offset along the time axis by adding 773 random jitter. (b) Map highlighting geographic areas (coloured areas) for samples included in the individual regional 774 timelines, and excavation locations (black crosses). Only shotgun sequenced genomes were used in our study, why the 775 exact timing of ancestry shifts may differ slightly from previous studies if based on different types of data from 776 different individuals. 777 Fig. 3: Spatiotemporal kriging analysis of major ancestries. The temporal transects demonstrate how WHG ancestry 778 (Italy_15000BP_9000BP), was replaced by Neolithic farmer ancestry (Boncuklu_10000BP) during the Neolithic 779 transition in Europe. Later, the steppe migrations around 5,000 cal. BP introduce both EHG (MiddleDon_7500BP) and 780 CHG (Caucasus_13000BP_10000BP) ancestry into Europe, thereby reducing Neolithic farmer ancestry. 781 Fig. 4: Fine-scale structure and temporal dynamics of steppe-related ancestry during the second transition in 782 Europe. (a) Correlation between estimated proportions of steppe-related and GAC farmer-related ancestries 783 (“postNeol” source set), across west Eurasian target individuals. (b) Timeline of difference in estimated steppe-related 784 ancestry proportions, using individuals from genetic cluster “Steppe_5000BP_4300BP” associated with either Yamnaya 785 or Afansievo cultural contexts as separate sources. Individuals from European post-Neolithic genetic clusters before 786 3,000 cal. BP are indicated with coloured symbols, while other west Eurasian target individuals are indicated with grey 787 symbols. Symbols with black outlines highlight early steppe-related individuals associated with either Corded Ware or 788 related (e.g., Battle Axe) cultural contexts. 789 790 Fig. 5: Genetic transects east of the Urals. Timelines of genetic ancestry compositions within the past 6,000 years 791 east of the Urals. Shown are ancestry proportions in 148 imputed ancient genomes from this region, inferred using 792 supervised ancestry modelling (“postNeol” source set). Panels separate ancestry proportions from local forest steppe 793 HGs (HG) and sources representing ancestries originating further east or west. 794 795 Fig. 6: Genetic relatedness across Western Eurasia. Maps showing networks of highest IBD sharing (top 10 highest 796 sharing per individual) during different time periods for 579 imputed genomes predating 3,000 cal. BP and located in 797 the geographic region shown. Shading and thickness of lines is scaled to represent the amount of IBD shared between 798 two individuals. In the earliest periods, sharing networks exhibit strong links within relatively narrow geographic 799 regions, representing predominantly close genetic ties between small HG communities, and rarely crossing the East-800 West divide extending from the Baltic to the Black Sea. From ~9,000 cal. BP onwards, a more extensive network with 801 weaker individual ties appears in the south, linking Anatolia to the rest of Europe, as early Neolithic farmer 802 communities spread across the continent. The period 7,000 - 5,000 cal. BP shows more connected subnetworks of 803 western European and eastern/northern European Neolithic farmers, while locally connected networks of HG 804 communities prevail on the eastern side of the divide. From c. 5,000 BP onwards the divide finally collapses, and 805 continental-wide genetic relatedness unifies large parts of Western Eurasia. 806 807 808 Methods 809 aDNA data generation and authentication 810 Sampling of ancient human remains was undertaken in collaboration with co-authors responsible 811 for the curation and contextual analyses of these, and with approval of the relevant institutions 812 responsible for the archaeological remains (detailed in the Reporting Summary). Laboratory work 813 was undertaken in dedicated aDNA cleanlab facilities (Globe Institute, University of Copenhagen) 814 following optimised aDNA protocols1,76 (Supplementary Note 1). Double-stranded blunt-end 815 libraries were constructed from the extracted DNA using NEBNext DNA Prep Master Mix Set 816 E6070 (New England Biolabs Inc.) and sequenced (80bp and 100bp single read) on Illumina HiSeq 817 2500 and 4000 platforms. Initial shallow shotgun-screening identified 317 of 962 ancient samples 818 https://paperpile.com/c/xoiDNc/WMvJ+JUiwg 18 with sufficient DNA preservation for deeper sequencing. Of these, 211 were teeth, 91 were petrous 819 bones, and 15 were sampled from long bones, ribs and cranial bones (Supplementary Data II). 820 Reads were mapped to the human reference genome build 37 and also to the mitochondrial genome 821 (rCRS) alone. Mapped reads were filtered for mapping quality 30 and sorted using Picard (v.1.127) 822 (http://picard.sourceforge.net) and samtools77. Data was merged to library level and duplicates 823 removed using Picard MarkDuplicates (v.1.127) and merged to sample level. Sample-level BAMs 824 were re-aligned using GATK (v.3.3.0) and hereafter had the md-tag updated and extended BAQs 825 calculated using samtools calmd (v.1.10)77. Read depth and coverage were determined using pysam 826 (https://github.com/pysam-developers/pysam) and BEDtools (v.2.23.0)78. Post-mortem DNA 827 damage patterns were determined using mapDamage2.079. For the 317 samples we observed C-to-T 828 deamination fractions ranging from 10.4% to 67.8%, with an average of 38.3% across all samples 829 (Supplementary Data I). These numbers indicate DNA molecule degradation consistent with a 830 millennia-scale depositional age. Three different methods were used to estimate DNA 831 contamination: two based on mitochondrial sequences80,81 and one method investigating X-832 chromosomal data in males (ANGSD, Supplementary Note 1). All contamination estimates are 833 reported in Supplementary Data V (summary values in Supplementary Data I). Based on this 834 approach we have a total of 15 samples flagged as ‘possibly contaminated’ in our downstream 835 analyses (Supplementary Note 1). 836 837 Imputation of ancient genomes 838 We imputed the ancient genomes in this study using the imputation and phasing tool GLIMPSE 839 v1.0.034 and 1000 Genomes phase335 as a reference panel. We first generated genotype likelihoods 840 at the biallelic 1000 Genomes variant sites from the bam files with bcftools v1.10 and the command 841 bcftools mpileup with parameters -I -E -a 'FORMAT/DP' --ignore-RG, followed by bcftools call -842 Aim -C alleles. Using GLIMPSE_chunk, the genotype likelihood data were first split into chunks of 843 sizes between 1 and 2 Mb with a buffer region of 200 kb at each side. We then imputed each chunk 844 with GLIMPSE_phase with parameters --burn 10, --main 15 and --pbwt-depth 2. Finally, the 845 imputed chunks were ligated with GLIMPSE_ligate. To validate the accuracy of the imputation, 42 846 high coverage (5X to 39X) genomes, including a Neolithic trio, were downsampled for testing82 847 (Supplementary Note 2). We evaluated imputation accuracy on the basis of depth of coverage; 848 minor allele frequency; and ancestry and timeframe of ancient genomes, using high coverage 849 ancient genomes82. >1X genomes provided remarkably high imputation accuracy (closely matching 850 that obtained for modern samples, Extended Data Fig. 2), except for African genomes that had 851 lower accuracy due to poor representation of this ancestry in the reference panel. Imputation 852 accuracy was influenced by both MAF and coverage (Supplementary Fig. S2.3). We found that 853 coverage as low as 0.1X and 0.4X was sufficient to obtain r2 imputation accuracy of 0.8 and 0.9 at 854 common variants (MAF≥10%), respectively. We conclude that ancient genomes can be imputed 855 confidently from coverages above 0.4X, and genome-wide aggregate analyses relying on common 856 SNPs (e.g. PCA and admixture modelling) can be carried out with a low amount of bias for genome 857 coverage from as low as 0.1X when using specific QC on the imputed data (although at very low 858 coverage a bias arises towards the major allele, see Supplementary Note 2). We additionally tested 859 for possible effects of bias affecting inferred ancestry components82 propagating biases in 860 individual-level pairwise analyses, using D-statistics, which indicated that imputed ancient genomes 861 down to 0.1x coverage are not significantly affected (Supplementary Note 2). 862 863 864 http://picard.sourceforge.net/ https://paperpile.com/c/xoiDNc/GcqeB https://paperpile.com/c/xoiDNc/GcqeB https://github.com/pysam-developers/pysam https://paperpile.com/c/xoiDNc/FSK6G https://paperpile.com/c/xoiDNc/a1yjs https://paperpile.com/c/xoiDNc/oCWP1+Jqrww https://paperpile.com/c/xoiDNc/4XCxT https://paperpile.com/c/xoiDNc/1NZld https://paperpile.com/c/xoiDNc/6kphW https://paperpile.com/c/xoiDNc/6kphW https://paperpile.com/c/xoiDNc/6kphW 19 Demographic inference 865 We determined the genetic sex of the study individuals using the ratio of reads aligning to either of 866 the sex chromosomes (RY statistic)83. Y chromosomes of inferred male individuals were further 867 analysed using phylogenetic placement84. We built a reference phylogenetic tree of 1,244 male 868 individuals from the 1000 Genomes project with RAxML-NG85, using the general time-reversible 869 model including among-site rate heterogeneity and ascertainment correction (model 870 GTR+G+ASC_LEWIS). For each ancient sample, haploid genotypes given the positions and alleles 871 in the reference panel were called using ‘bcftools call’ (options -C alleles –ploidy 1 -i). The 872 resulting genotypes were converted to fasta format and placed onto the reference tree using EPA-873 ng84. Phylogenetic placements were processed and visualised using gappa86. To convert 874 phylogenetic placements into haplogroup calls, we assigned each branch of the reference phylogeny 875 to its representing haplogroup, using SNP annotations from ISOGG (version 15.73). For each 876 ancient sample, haplogroups were then called using the most basal branch accumulating 99% of the 877 placement weights, obtained using accumulate in gappa. Phylogenetic analyses of reconstructed 878 mitochondrial genomes were also undertaken using RAxML-ng84 (see Supplementary Note 3a). 879 To infer genetic relatedness between the study individuals we used the allele-frequency free 880 inference method introduced by87. For each pair of individuals, three relatedness estimators were 881 calculated, R0, R1 and KING-robust88 using the site-frequency-spectrum (SFS)-based approach. 882 We used the realSFS89 method implemented in the ANGSD90 package to infer the 2D-SFS, 883 selecting the SFS with the highest likelihood across ten replicates. We used a set of 1,191,529 884 autosomal transversion SNPs with minor allele frequency ≥ 0.05 from the 1000 Genomes Project35 885 for the analysis. Previously established cut-offs88 for the KING-robust estimator were applied to 886 assign individual pairs to first-, second- or third-degree relationships. Parent-offspring relationships 887 were distinguished from sibling relationships using R0 and R1 ratios, by requiring that R0 ≤ 0.02 888 and 0.4 ≤ R1 ≤ 0.6 to infer a parent-offspring relative pair. Individual pairs with less than 20,000 889 sites contributing to the estimators were excluded. 890 We generated a dataset for population genetic analysis by combining the 317 newly sequenced 891 individuals with 1,347 previously published ancient genomes with genomic coverage >0.1X 892 generated using shotgun sequencing (Supplementary Data VII). Imputed genotype data 893 (Supplementary Note 2) for this set of 1,664 ancient genomes were merged with genotypes of 2,504 894 modern individuals from the 1,000 Genomes project35 used as a reference panel in the imputation. 895 We retained only SNPs passing the 1000 Genomes strict mask, resulting in a final dataset of 4,168 896 individuals genotyped at 7,321,965 autosomal SNPs (“1000G” dataset). In addition to imputed 897 genotypes, we also generated pseudo-haploid genotypes for each ancient individual by randomly 898 sampling an allele from sequencing reads covering those SNPs. For population structure analyses in 899 the context of global genetic diversity, we generated a second dataset by intersecting the ancient 900 genotype data with SNP array data of 2,180 modern individuals from 213 world-wide 901 populations3,4,91,92 (“HO” dataset). 902 To facilitate filtering for downstream analyses, we flagged individuals to potentially exclude based 903 on the following criteria: i) Contamination estimate >5% (“contMT5pct”, “contNuc5pct”; 904 Supplementary Note 1); ii) Autosomal coverage <0.1X (“lowcov”), III) Genome-wide average 905 imputation genotype probability <0.98 (“lowGpAvg”), IV) Individual is the lower quality sample in 906 a close relative pair (“1d_rel”, “2d_rel”; Supplementary Note 3c). A total of 1,492 individuals (213 907 newly reported) passed all filters, which were used in the majority of downstream analyses unless 908 otherwise noted. 909 We investigated overall population structure among the dataset individuals using principal 910 component analyses (PCA) and model-based clustering (ADMIXTURE93). We carried out PCA 911 using different subsets of individuals in the “HO'' dataset. For the PCA including only imputed 912 diploid samples, we used GCTA94, excluding SNPs with minor allele frequency (MAF) < 0.05 in 913 https://paperpile.com/c/xoiDNc/KvO9F https://paperpile.com/c/xoiDNc/CuC5C https://paperpile.com/c/xoiDNc/gtBem https://paperpile.com/c/xoiDNc/CuC5C https://paperpile.com/c/xoiDNc/jmJu0 https://paperpile.com/c/xoiDNc/CuC5C https://paperpile.com/c/xoiDNc/1twkZ https://paperpile.com/c/xoiDNc/5cdUq https://paperpile.com/c/xoiDNc/5m7Pf https://paperpile.com/c/xoiDNc/PoK3m https://paperpile.com/c/xoiDNc/1NZld https://paperpile.com/c/xoiDNc/5cdUq https://paperpile.com/c/xoiDNc/1NZld https://paperpile.com/c/xoiDNc/tXEcL+bPGT+etCF+Afv5S https://paperpile.com/c/xoiDNc/skNv https://paperpile.com/c/xoiDNc/ItIpZ 20 the respective panel. For PCA projecting low coverage or flagged individuals, we used smartpca95,96 914 with options ‘lsqproject: YES’ and ‘autoshrink: YES’ on a fixed set of 400,186 SNPs with MAF ≥ 915 0.05 in non-African individuals passing all filters. We ran ADMIXTURE on a set of 1,593 ancient 916 individuals from the “1000G” dataset, excluding individuals flagged as close relatives or a 917 contamination estimate >5%. For the 1,492 individuals passing all filters we used imputed 918 genotypes, the remaining 101 lower coverage samples were represented by pseudo-haploid 919 genotypes. We restricted the analysis to transversion SNPs with imputation INFO score ≥ 0.8 and 920 MAF ≥ 0.05. We further performed linkage disequilibrium (LD) pruning and filtering for 921 missingness using plink97 (options --indep-pairwise 500 50 0.4 –geno 0.8), for a final analysis set of 922 142,550 SNPs. 923 We performed admixture graph fitting (qpGraph) to investigate deep Eurasian population structure 924 using ADMIXTOOLS298. For these analyses, pairwise f2-statistics were pre-computed from 925 pseudo-haploid genotypes in the “1000G” dataset using the ‘extract_f2’ function with 926 ‘afProd=TRUE’. We grouped individuals into populations using their membership in the genetic 927 clusters inferred from IBD sharing (Supplementary Note 3f), with the exception of the UP European 928 individual Kostenki 14, which was treated as a separate population (new cluster label 929 “Europe_37000BP_33000BP_Kostenki”). We carried out admixture graph fitting using a semi-930 automatic iterative approach (Supplementary Note 3d). 931 We used IBDseq99 to detect genomic segments shared identical-by-descent (IBD) between all 932 individuals in the “1000G” dataset, restricting to transversion SNPs with imputation INFO score ≥ 933 0.8 and MAF ≥ 0.01. We filtered the resulting IBD segments for LOD score ≥ 3 and a minimum 934 length of 2 centimorgans (cM), and further removed regions of excess long IBD following100. First, 935 we used the GenomicRanges101 package in R to calculate the total number of long IBD segments 936 (>10cm) overlapping each position along the genome, and calculated their 3% trimmed mean and 937 standard deviation (SD). We then called regions of excess IBD if they were > 10 trimmed SD from 938 the trimmed mean, and removed any segments overlapping the excess IBD regions. For analyses of 939 runs of homozygosity (ROH) we used a shorter length cutoff of 1cM. 940 We carried out genetic clustering of the ancient individuals using hierarchical community detection 941 on a network of pairwise identity-by-descent (IBD)-sharing similarities102. To facilitate detection of 942 clusters at a finer scale, we ran IBDseq (version r1206) on a dataset restricting to ancient samples 943 only, and applied more lenient filters of imputation INFO score > 0.5, and minimum IBD segment 944 length of 1 cM. We constructed a weighted network of the individuals using the igraph103 package 945 in R, with the fraction of the genome shared IBD between pairs of individuals as weights. We then 946 performed iterative community detection on this network using the Leiden algorithm104 947 implemented in the leidenAlg R package (v1.01, https://github.com/kharchenkolab/leidenAlg). We 948 used a resolution parameter of r=0.5 as the starting value for each level of community detection. If 949 more than one community was detected, we split the network into the respective communities, and 950 repeated the community detection step. If no communities were detected, we incremented the 951 resolution parameter in steps of 0.5 until a maximum value of r=3. The initial clustering was 952 completed when no more communities were detected at the highest resolution parameter, across all 953 subcommunities. To convert the resulting hierarchy into a final clustering, we simplified the initial 954 clustering by collapsing nodes into single clusters based on observed spatiotemporal annotations of 955 the samples. We note that the obtained clusters should not be interpreted as ‘populations’ in the 956 sense of a local community of individuals, but rather as sets of individuals with shared ancestry. 957 While this approach is an oversimplification of the complex spatiotemporally structured populations 958 investigated here, the obtained clusters nevertheless captured real effects, grouping individuals 959 within restricted spatiotemporal ranges and/or archaeological contexts and recapitulating known 960 relationships between clusters. 961 962 https://paperpile.com/c/xoiDNc/msOlR+DRR9h https://paperpile.com/c/xoiDNc/IV5vS https://paperpile.com/c/xoiDNc/H8S8I https://paperpile.com/c/xoiDNc/ESPFs https://paperpile.com/c/xoiDNc/sAFax https://paperpile.com/c/xoiDNc/0QvHH https://paperpile.com/c/xoiDNc/k9UIQ https://paperpile.com/c/xoiDNc/Fpg6i https://paperpile.com/c/xoiDNc/LG1Er 21 To circumvent some of the pitfalls of grouping individuals into discrete clusters, we used 963 supervised ancestry modelling where sets of ‘target’ individuals were modelled as mixtures of 964 ‘source’ groups, selected to represent particular ancestry components. As an illustrative case, an 965 individual of European HG ancestry with a minor contribution of Neolithic farmer admixture might 966 be inferred to be a member of a HG genetic cluster, but will be modelled as a mixture of a HG and 967 Neolithic farmer sources in the ancestry modelling. To estimate ancestry proportions from patterns 968 of pairwise IBD sharing, we applied an approach akin to “chromosome painting”105. We first 969 inferred an IBD-based “painting profile” for each target individual, by summing up the total amount 970 of IBD shared with each “donor” group (using population labels for modern donors or IBD-based 971 genetic clusters for ancient donors), and normalising them to the interval [0,1]. We used a leave-972 one-out approach following37 to account for the fact that recipient individuals cannot be included as 973 donors from their own group. We then used these painting profiles in supervised modelling of target 974 individuals as mixtures from different sets of putative source groups37,106, using non-negative least 975 squares implemented in the R package limSolve107. We estimated standard errors of ancestry 976 proportions using a weighted block jacknife, leaving out each chromosome in turns. A comparison 977 of results obtained using this approach to other commonly used methods (supervised 978 ADMIXTURE, qpAdm) is shown in Supplementary Note 3f). We focussed our analyses on three 979 panels of putative source clusters reflecting different temporal depths: ‘deep’, using a set of deep 980 ancestry source groups reflecting major ancestry poles; ‘postNeol’, using diverse Neolithic and 981 earlier source groups; and ‘postBA’, using Late Neolithic and Bronze Age source groups (Extended 982 Data Figs. 5-7). We also used additional source sets in follow-up analyses of more restricted 983 spatiotemporal contexts (Supplementary Datas VII-XIII). 984 985 Finally, we aimed to infer the geographic and temporal spread of major ancestries (Supplementary 986 Note 3e). We used a method45 applying spatiotemporal ordinary kriging on latent ancestry 987 proportion estimates from ancient and present-day genomes. This way, we obtained spatiotemporal 988 maps reflecting the dynamics of the spread of ancestry during the transition from the Mesolithic to 989 the Neolithic, Bronze Age, Iron Age and more recent periods. We obtained ancestry proportions 990 estimated using ADMIXTURE108 with K=9 latent ancestry clusters (Supplementary Note 3d) on a 991 sequence dataset including both whole-genome shotgun-sequenced genomes and genomic 992 sequences obtained via SNP capture (Supplementary Note 2, intersection with “HO” dataset). We 993 performed spatiotemporal kriging109 of these proportions over the last 12,900 years, in intervals of 994 300 years, with a 5,000-point spatial grid spanning Western and Central Eurasia. We used the R 995 package gstat to fit a spatiotemporal variogram via a metric covariance model, and perform 996 ordinary kriging110. We focused on the ancestry clusters for which we could fit variogram models 997 that were not static over time. 998 999 14C chronology and reservoir effects 1000 Of the 317 individuals sequenced in this study 272 were 14C-dated in the project, while 30 14C-dates 1001 were obtained from literature, and 15 were dated by archaeological context (Supplementary Note 4, 1002 Supplementary Data II). Some individuals were dated twice. Most of the dates (n=242) were 1003 performed at the 14CHRONO Centre laboratory at Queen’s University, Belfast, following published 1004 sample pretreatment and laboratory protocols111. Additional samples were analysed by the Oxford 1005 Radiocarbon Accelerator Unit (ORAU) laboratory (n=24) and by the Keck-CCAMS Group, Irvine, 1006 California, USA (n=6) (see 112,113 for laboratory procedures). Only datings with a C/N ratio of 2.9-1007 3.6 were accepted; both δ13C and δ15N collagen measurements were also performed, and were 1008 https://paperpile.com/c/xoiDNc/n8XOg https://paperpile.com/c/xoiDNc/C9mFL https://paperpile.com/c/xoiDNc/C9mFL+QaZbo https://paperpile.com/c/xoiDNc/hm9Vq https://paperpile.com/c/xoiDNc/Qz5B6 https://paperpile.com/c/xoiDNc/X4Wpc https://paperpile.com/c/xoiDNc/P6xJY https://paperpile.com/c/xoiDNc/pzrYf https://paperpile.com/c/xoiDNc/jlmHf/?noauthor=1 https://paperpile.com/c/xoiDNc/AtPVS+Y0rcQ 22 used in estimates of marine (MRE) and freshwater (FRE) reservoir effects (see Supplementary Note 1009 4, Supplementary Data IV). Published values of MRE and FRE were used where available, but for 1010 some regions, such as sites in western Russia, a standard FRE value of 500 years was applied. A 1011 diet-weighted reservoir offset was then applied to the 14C central value before calibration. 1012 Calibrations were made in Oxcal 4.4 using the Intcal20 calibration curve114. For display and 1013 calculation purposes a midpoint of the reservoir corrected and calibrated 95% interval was 1014 calculated. Full details of the reservoir correction and calibration procedure are given in 1015 Supplementary Note 4 and the calculations are found in Table S4.1. 1016 1017 References (Methods) 1018 76. Damgaard, P. B. et al. Improving access to endogenous DNA in ancient bones and teeth. Sci. Rep. 5, 11184 1019 (2015). 1020 77. Li, H. et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics 25, 2078–2079 (2009). 1021 78. Quinlan, A. R. & Hall, I. M. BEDTools: a flexible suite of utilities for comparing genomic features. 1022 Bioinformatics 26, 841–842 (2010). 1023 79. Jónsson, H., Ginolhac, A., Schubert, M., Johnson, P. L. F. & Orlando, L. mapDamage2.0: fast approximate 1024 Bayesian estimates of ancient DNA damage parameters. Bioinformatics 29, 1682–1684 (2013). 1025 80. Renaud, G., Slon, V., Duggan, A. T. & Kelso, J. Schmutzi: estimation of contamination and endogenous 1026 mitochondrial consensus calling for ancient DNA. Genome Biol. 16, 224 (2015). 1027 81. Fu, Q. et al. Genome sequence of a 45,000-year-old modern human from western Siberia. Nature 514, 445–1028 449 (2014). 1029 82. Sousa da Mota, B. et al. Imputation of ancient human genomes. Nat. Commun. 14, 3660 (2023). 1030 83. Skoglund, P., Storå, J., Götherström, A. & Jakobsson, M. Accurate sex identification of ancient human 1031 remains using DNA shotgun sequencing. J. Archaeol. Sci. 40, 4477–4482 (2013). 1032 84. Barbera, P. et al. EPA-ng: Massively Parallel Evolutionary Placement of Genetic Sequences. Syst. Biol. 68, 1033 365–369 (2019). 1034 85. Kozlov, A. M., Darriba, D., Flouri, T., Morel, B. & Stamatakis, A. RAxML-NG: a fast, scalable and user-1035 friendly tool for maximum likelihood phylogenetic inference. Bioinformatics 35, 4453–4455 (2019). 1036 86. Czech, L., Barbera, P. & Stamatakis, A. Genesis and Gappa: processing, analyzing and visualizing 1037 phylogenetic (placement) data. Bioinformatics 36, 3263–3265 (2020). 1038 87. Waples, R. K., Albrechtsen, A. & Moltke, I. Allele frequency-free inference of close familial relationships 1039 from genotypes or low-depth sequencing data. Mol. Ecol. 28, 35–48 (2019). 1040 88. Manichaikul, A. et al. Robust relationship inference in genome-wide association studies. Bioinformatics 26, 1041 2867–2873 (2010). 1042 89. Nielsen, R., Korneliussen, T., Albrechtsen, A., Li, Y. & Wang, J. SNP calling, genotype calling, and sample 1043 allele frequency estimation from New-Generation Sequencing data. PLoS One 7, e37558 (2012). 1044 90. Korneliussen, T. S., Albrechtsen, A. & Nielsen, R. ANGSD: Analysis of Next Generation Sequencing Data. 1045 BMC Bioinformatics 15, 356 (2014). 1046 91. Patterson, N. et al. Ancient admixture in human history. Genetics 192, 1065–1093 (2012). 1047 92. Pickrell, J. K. et al. The genetic prehistory of southern Africa. Nat. Commun. 3, 1143 (2012). 1048 93. Shringarpure, S. S., Bustamante, C. D., Lange, K. & Alexander, D. H. Efficient analysis of large datasets and 1049 https://paperpile.com/c/xoiDNc/G0q5l http://paperpile.com/b/xoiDNc/JUiwg http://paperpile.com/b/xoiDNc/JUiwg http://paperpile.com/b/xoiDNc/JUiwg http://paperpile.com/b/xoiDNc/JUiwg http://paperpile.com/b/xoiDNc/JUiwg http://paperpile.com/b/xoiDNc/JUiwg http://paperpile.com/b/xoiDNc/JUiwg http://paperpile.com/b/xoiDNc/JUiwg http://paperpile.com/b/xoiDNc/GcqeB http://paperpile.com/b/xoiDNc/GcqeB http://paperpile.com/b/xoiDNc/GcqeB http://paperpile.com/b/xoiDNc/GcqeB http://paperpile.com/b/xoiDNc/GcqeB http://paperpile.com/b/xoiDNc/GcqeB http://paperpile.com/b/xoiDNc/GcqeB http://paperpile.com/b/xoiDNc/FSK6G http://paperpile.com/b/xoiDNc/FSK6G http://paperpile.com/b/xoiDNc/FSK6G http://paperpile.com/b/xoiDNc/FSK6G http://paperpile.com/b/xoiDNc/FSK6G http://paperpile.com/b/xoiDNc/FSK6G http://paperpile.com/b/xoiDNc/a1yjs http://paperpile.com/b/xoiDNc/a1yjs http://paperpile.com/b/xoiDNc/a1yjs http://paperpile.com/b/xoiDNc/a1yjs http://paperpile.com/b/xoiDNc/a1yjs http://paperpile.com/b/xoiDNc/a1yjs http://paperpile.com/b/xoiDNc/oCWP1 http://paperpile.com/b/xoiDNc/oCWP1 http://paperpile.com/b/xoiDNc/oCWP1 http://paperpile.com/b/xoiDNc/oCWP1 http://paperpile.com/b/xoiDNc/oCWP1 http://paperpile.com/b/xoiDNc/oCWP1 http://paperpile.com/b/xoiDNc/Jqrww http://paperpile.com/b/xoiDNc/Jqrww http://paperpile.com/b/xoiDNc/Jqrww http://paperpile.com/b/xoiDNc/Jqrww http://paperpile.com/b/xoiDNc/Jqrww http://paperpile.com/b/xoiDNc/Jqrww http://paperpile.com/b/xoiDNc/Jqrww http://paperpile.com/b/xoiDNc/Jqrww http://paperpile.com/b/xoiDNc/6kphW http://paperpile.com/b/xoiDNc/6kphW http://paperpile.com/b/xoiDNc/6kphW http://paperpile.com/b/xoiDNc/6kphW http://paperpile.com/b/xoiDNc/6kphW http://paperpile.com/b/xoiDNc/6kphW http://paperpile.com/b/xoiDNc/6kphW http://paperpile.com/b/xoiDNc/KvO9F http://paperpile.com/b/xoiDNc/KvO9F http://paperpile.com/b/xoiDNc/KvO9F http://paperpile.com/b/xoiDNc/KvO9F http://paperpile.com/b/xoiDNc/KvO9F http://paperpile.com/b/xoiDNc/KvO9F http://paperpile.com/b/xoiDNc/CuC5C http://paperpile.com/b/xoiDNc/CuC5C http://paperpile.com/b/xoiDNc/CuC5C http://paperpile.com/b/xoiDNc/CuC5C http://paperpile.com/b/xoiDNc/CuC5C http://paperpile.com/b/xoiDNc/CuC5C http://paperpile.com/b/xoiDNc/CuC5C http://paperpile.com/b/xoiDNc/CuC5C http://paperpile.com/b/xoiDNc/gtBem http://paperpile.com/b/xoiDNc/gtBem http://paperpile.com/b/xoiDNc/gtBem http://paperpile.com/b/xoiDNc/gtBem http://paperpile.com/b/xoiDNc/gtBem http://paperpile.com/b/xoiDNc/gtBem http://paperpile.com/b/xoiDNc/jmJu0 http://paperpile.com/b/xoiDNc/jmJu0 http://paperpile.com/b/xoiDNc/jmJu0 http://paperpile.com/b/xoiDNc/jmJu0 http://paperpile.com/b/xoiDNc/jmJu0 http://paperpile.com/b/xoiDNc/jmJu0 http://paperpile.com/b/xoiDNc/1twkZ http://paperpile.com/b/xoiDNc/1twkZ http://paperpile.com/b/xoiDNc/1twkZ http://paperpile.com/b/xoiDNc/1twkZ http://paperpile.com/b/xoiDNc/1twkZ http://paperpile.com/b/xoiDNc/1twkZ http://paperpile.com/b/xoiDNc/5cdUq http://paperpile.com/b/xoiDNc/5cdUq http://paperpile.com/b/xoiDNc/5cdUq http://paperpile.com/b/xoiDNc/5cdUq http://paperpile.com/b/xoiDNc/5cdUq http://paperpile.com/b/xoiDNc/5cdUq http://paperpile.com/b/xoiDNc/5cdUq http://paperpile.com/b/xoiDNc/5cdUq http://paperpile.com/b/xoiDNc/5m7Pf http://paperpile.com/b/xoiDNc/5m7Pf http://paperpile.com/b/xoiDNc/5m7Pf http://paperpile.com/b/xoiDNc/5m7Pf http://paperpile.com/b/xoiDNc/5m7Pf http://paperpile.com/b/xoiDNc/5m7Pf http://paperpile.com/b/xoiDNc/PoK3m http://paperpile.com/b/xoiDNc/PoK3m http://paperpile.com/b/xoiDNc/PoK3m http://paperpile.com/b/xoiDNc/PoK3m http://paperpile.com/b/xoiDNc/PoK3m http://paperpile.com/b/xoiDNc/PoK3m http://paperpile.com/b/xoiDNc/tXEcL http://paperpile.com/b/xoiDNc/tXEcL http://paperpile.com/b/xoiDNc/tXEcL http://paperpile.com/b/xoiDNc/tXEcL http://paperpile.com/b/xoiDNc/tXEcL http://paperpile.com/b/xoiDNc/tXEcL http://paperpile.com/b/xoiDNc/tXEcL http://paperpile.com/b/xoiDNc/Afv5S http://paperpile.com/b/xoiDNc/Afv5S http://paperpile.com/b/xoiDNc/Afv5S http://paperpile.com/b/xoiDNc/Afv5S http://paperpile.com/b/xoiDNc/Afv5S http://paperpile.com/b/xoiDNc/Afv5S http://paperpile.com/b/xoiDNc/Afv5S http://paperpile.com/b/xoiDNc/skNv 23 sex bias with ADMIXTURE. BMC Bioinformatics 17, 218 (2016). 1050 94. Yang, J., Lee, S. H., Goddard, M. E. & Visscher, P. M. GCTA: a tool for genome-wide complex trait analysis. 1051 The American Journal of (2011). 1052 95. Patterson, N., Price, A. L. & Reich, D. Population structure and eigenanalysis. PLoS Genet. 2, e190 (2006). 1053 96. Price, A. L. et al. Principal components analysis corrects for stratification in genome-wide association studies. 1054 Nat. Genet. 38, 904–909 (2006). 1055 97. Chang, C. C. et al. Second-generation PLINK: rising to the challenge of larger and richer datasets. 1056 Gigascience 4, 7 (2015). 1057 98. Maier, R., Flegontov, P., Flegontova, O., Changmai, P. & Reich, D. On the limits of fitting complex models of 1058 population history to genetic data. bioRxiv 2022.05.08.491072 (2022) doi:10.1101/2022.05.08.491072. 1059 99. Browning, B. L. & Browning, S. R. Detecting identity by descent and estimating genotype error rates in 1060 sequence data. Am. J. Hum. Genet. 93, 840–851 (2013). 1061 100. Browning, S. R. & Browning, B. L. Accurate Non-parametric Estimation of Recent Effective 1062 Population Size from Segments of Identity by Descent. Am. J. Hum. Genet. 97, 404–418 (2015). 1063 101. Lawrence, M. et al. Software for computing and annotating genomic ranges. PLoS Comput. Biol. 9, 1064 e1003118 (2013). 1065 102. Greenbaum, G., Rubin, A., Templeton, A. R. & Rosenberg, N. A. Network-based hierarchical 1066 population structure analysis for large genomic data sets. Genome Res. 29, 2020–2033 (2019). 1067 103. Csardi, G., Nepusz, T. & Others. The igraph software package for complex network research. 1068 InterJournal, complex systems 1695, 1–9 (2006). 1069 104. Traag, V. A., Waltman, L. & van Eck, N. J. From Louvain to Leiden: guaranteeing well-connected 1070 communities. Sci. Rep. 9, 5233 (2019). 1071 105. Lawson, D. J., Hellenthal, G., Myers, S. & Falush, D. Inference of population structure using dense 1072 haplotype data. PLoS Genet. 8, e1002453 (2012). 1073 106. Hellenthal, G. et al. A genetic atlas of human admixture history. Science 343, 747–751 (2014). 1074 107. Soetaert, K., Van den Meersche, K. & van Oevelen, D. limSolve: Solving linear inverse models. 1075 (2009). 1076 108. Alexander, D. H., Novembre, J. & Lange, K. Fast model-based estimation of ancestry in unrelated 1077 individuals. Genome Res. 19, 1655–1664 (2009). 1078 109. Cressie, N. & Wikle, C. K. Statistics for Spatio-Temporal Data. (John Wiley & Sons, 2015). 1079 110. Gräler, B., Pebesma, E. & Heuvelink, G. Spatio-Temporal Interpolation using gstat. R J. 8, 204 1080 (2016). 1081 111. Reimer, P., Hoper, S., MacDonald, J., Reimer, R. & Thompson, M. Laboratory Protocols used for 1082 AMS Radiocarbon Dating at the 14CHRONO Centre. (2015). 1083 112. Brock, F., Higham, T., Ditchfield, P. & Ramsey, C. B. Current Pretreatment Methods for AMS 1084 Radiocarbon Dating at the Oxford Radiocarbon Accelerator Unit (Orau. vol. Radiocarbon52 103–112 Preprint at 1085 (2010). 1086 113. Beaumont, W., Beverly, R., Southon, J. & Taylor, R. E. Bone preparation at the KCCAMS 1087 laboratory. Nucl. Instrum. Methods Phys. Res. B 268, 906–909 (2010). 1088 114. Reimer, P., Austin, W. & Bard, E. The IntCal20 Northern Hemisphere Radiocarbon Age Calibration 1089 http://paperpile.com/b/xoiDNc/skNv http://paperpile.com/b/xoiDNc/skNv http://paperpile.com/b/xoiDNc/skNv http://paperpile.com/b/xoiDNc/skNv http://paperpile.com/b/xoiDNc/skNv http://paperpile.com/b/xoiDNc/ItIpZ http://paperpile.com/b/xoiDNc/ItIpZ http://paperpile.com/b/xoiDNc/ItIpZ http://paperpile.com/b/xoiDNc/ItIpZ http://paperpile.com/b/xoiDNc/msOlR http://paperpile.com/b/xoiDNc/msOlR http://paperpile.com/b/xoiDNc/msOlR http://paperpile.com/b/xoiDNc/msOlR http://paperpile.com/b/xoiDNc/msOlR http://paperpile.com/b/xoiDNc/DRR9h http://paperpile.com/b/xoiDNc/DRR9h http://paperpile.com/b/xoiDNc/DRR9h http://paperpile.com/b/xoiDNc/DRR9h http://paperpile.com/b/xoiDNc/DRR9h http://paperpile.com/b/xoiDNc/DRR9h http://paperpile.com/b/xoiDNc/DRR9h http://paperpile.com/b/xoiDNc/DRR9h http://paperpile.com/b/xoiDNc/IV5vS http://paperpile.com/b/xoiDNc/IV5vS http://paperpile.com/b/xoiDNc/IV5vS http://paperpile.com/b/xoiDNc/IV5vS http://paperpile.com/b/xoiDNc/IV5vS http://paperpile.com/b/xoiDNc/IV5vS http://paperpile.com/b/xoiDNc/IV5vS http://paperpile.com/b/xoiDNc/IV5vS http://paperpile.com/b/xoiDNc/H8S8I http://paperpile.com/b/xoiDNc/H8S8I http://paperpile.com/b/xoiDNc/H8S8I http://paperpile.com/b/xoiDNc/H8S8I http://dx.doi.org/10.1101/2022.05.08.491072 http://dx.doi.org/10.1101/2022.05.08.491072 http://paperpile.com/b/xoiDNc/ESPFs http://paperpile.com/b/xoiDNc/ESPFs http://paperpile.com/b/xoiDNc/ESPFs http://paperpile.com/b/xoiDNc/ESPFs http://paperpile.com/b/xoiDNc/ESPFs http://paperpile.com/b/xoiDNc/ESPFs http://paperpile.com/b/xoiDNc/sAFax http://paperpile.com/b/xoiDNc/sAFax http://paperpile.com/b/xoiDNc/sAFax http://paperpile.com/b/xoiDNc/sAFax http://paperpile.com/b/xoiDNc/sAFax http://paperpile.com/b/xoiDNc/sAFax http://paperpile.com/b/xoiDNc/0QvHH http://paperpile.com/b/xoiDNc/0QvHH http://paperpile.com/b/xoiDNc/0QvHH http://paperpile.com/b/xoiDNc/0QvHH http://paperpile.com/b/xoiDNc/0QvHH http://paperpile.com/b/xoiDNc/0QvHH http://paperpile.com/b/xoiDNc/0QvHH http://paperpile.com/b/xoiDNc/0QvHH http://paperpile.com/b/xoiDNc/k9UIQ http://paperpile.com/b/xoiDNc/k9UIQ http://paperpile.com/b/xoiDNc/k9UIQ http://paperpile.com/b/xoiDNc/k9UIQ http://paperpile.com/b/xoiDNc/k9UIQ http://paperpile.com/b/xoiDNc/k9UIQ http://paperpile.com/b/xoiDNc/Fpg6i http://paperpile.com/b/xoiDNc/Fpg6i http://paperpile.com/b/xoiDNc/Fpg6i http://paperpile.com/b/xoiDNc/Fpg6i http://paperpile.com/b/xoiDNc/Fpg6i http://paperpile.com/b/xoiDNc/Fpg6i http://paperpile.com/b/xoiDNc/LG1Er http://paperpile.com/b/xoiDNc/LG1Er http://paperpile.com/b/xoiDNc/LG1Er http://paperpile.com/b/xoiDNc/LG1Er http://paperpile.com/b/xoiDNc/LG1Er http://paperpile.com/b/xoiDNc/LG1Er http://paperpile.com/b/xoiDNc/n8XOg http://paperpile.com/b/xoiDNc/n8XOg http://paperpile.com/b/xoiDNc/n8XOg http://paperpile.com/b/xoiDNc/n8XOg http://paperpile.com/b/xoiDNc/n8XOg http://paperpile.com/b/xoiDNc/n8XOg http://paperpile.com/b/xoiDNc/QaZbo http://paperpile.com/b/xoiDNc/QaZbo http://paperpile.com/b/xoiDNc/QaZbo http://paperpile.com/b/xoiDNc/QaZbo http://paperpile.com/b/xoiDNc/QaZbo http://paperpile.com/b/xoiDNc/QaZbo http://paperpile.com/b/xoiDNc/QaZbo http://paperpile.com/b/xoiDNc/hm9Vq http://paperpile.com/b/xoiDNc/hm9Vq http://paperpile.com/b/xoiDNc/hm9Vq http://paperpile.com/b/xoiDNc/hm9Vq http://paperpile.com/b/xoiDNc/X4Wpc http://paperpile.com/b/xoiDNc/X4Wpc http://paperpile.com/b/xoiDNc/X4Wpc http://paperpile.com/b/xoiDNc/X4Wpc http://paperpile.com/b/xoiDNc/X4Wpc http://paperpile.com/b/xoiDNc/X4Wpc http://paperpile.com/b/xoiDNc/P6xJY http://paperpile.com/b/xoiDNc/P6xJY http://paperpile.com/b/xoiDNc/P6xJY http://paperpile.com/b/xoiDNc/pzrYf http://paperpile.com/b/xoiDNc/pzrYf http://paperpile.com/b/xoiDNc/pzrYf http://paperpile.com/b/xoiDNc/pzrYf http://paperpile.com/b/xoiDNc/pzrYf http://paperpile.com/b/xoiDNc/pzrYf http://paperpile.com/b/xoiDNc/jlmHf http://paperpile.com/b/xoiDNc/jlmHf http://paperpile.com/b/xoiDNc/jlmHf http://paperpile.com/b/xoiDNc/jlmHf http://paperpile.com/b/xoiDNc/AtPVS http://paperpile.com/b/xoiDNc/AtPVS http://paperpile.com/b/xoiDNc/AtPVS http://paperpile.com/b/xoiDNc/Y0rcQ http://paperpile.com/b/xoiDNc/Y0rcQ http://paperpile.com/b/xoiDNc/Y0rcQ http://paperpile.com/b/xoiDNc/Y0rcQ http://paperpile.com/b/xoiDNc/Y0rcQ http://paperpile.com/b/xoiDNc/Y0rcQ http://paperpile.com/b/xoiDNc/G0q5l 24 Curve (0–55 cal kBP. Radiocarbon 62, 725–757 (2020). 1090 Extended Data Figures 1091 Extended Data Fig. 1: Genetic structure of the 317 herein-reported ancient genomes. (a-d) Principal component 1092 analysis of 3,316 modern and ancient individuals from Eurasia, Oceania and the Americas (a, b), as well as restricted to 1093 2,126 individuals from western Eurasia (west of the Urals) (c, d). Shown are analyses with principal components 1094 inferred either using both modern and imputed ancient genomes passing all filters, and projecting low coverage ancient 1095 genomes (a, c); or only modern genomes and projecting all ancient genomes (b, d). Ancient genomes sequenced in this 1096 study are indicated either with black circles (imputed genomes) or grey diamonds (projected genomes). (e) Model-based 1097 clustering results using ADMIXTURE for 284 newly reported genomes (excluding close relatives and individuals 1098 flagged for possible contamination ). Results shown are based on ADMIXTURE runs from K=2 to K=15 on 1,593 1099 ancient individuals, corresponding to the full set of 1,492 imputed genomes passing filters as well as 101 low coverage 1100 genomes represented by pseudo-haploid genotypes (flags “lowcov” or “lowGpAvg”, Supplementary Data VII; 1101 indicated with alpha transparency in plot). 1102 1103 Extended Data Fig. 2. Imputation accuracy of aDNA. (a) imputation accuracy across 42 high-coverage ancient 1104 genomes when downsampled to lower depth of coverage values (see Supplementary Note 2, Table 2.1). (b-d) 1105 imputation accuracy for 1X depth of coverage across 21 ancient European genomes. In all panels, imputation accuracy 1106 is shown as the squared Pearson correlation between imputed and true genotype dosages as a function of minor allele 1107 frequency of the target variant sites. 1108 1109 Extended Data Fig. 3. Genetic clustering of ancient individuals. Characterization of genetic clusters for 1,401 1110 imputed ancient individuals from Eurasia (i.e. excluding 91 individuals from Africa and Americas), inferred from 1111 pairwise identity-by-descent (IBD) sharing (indicated using coloured symbols throughout) (a) Temporal distribution of 1112 clustered individuals, grouped by broad ancestry cluster. (b), (c) Geographical distribution of clustered individuals, 1113 shown for individuals predating 3,000 BP (b) and after 3,000 BP (c). (d) Network graph of pairwise IBD sharing 1114 between 596 ancient Eurasians predating 3,000 BP, highlighting within- and between-cluster relationships. Each node 1115 represents an individual, and the width of edges connecting nodes indicates the fraction of the genome shared IBD 1116 between the respective pair of individuals. Network edges were restricted to the 10 highest sharing connections for each 1117 individual, and the layout was computed using the force-directed Fruchterman-Reingold algorithm. (e) Neighbour-1118 joining tree showing relationships between genetic clusters, inferred using total variation distance (TVD) of IBD 1119 painting palettes. (f), (g) PCA of 3,119 Eurasian (f) or 2,126 west Eurasian (g) ancient and modern individuals (“HO” 1120 dataset). 1121 1122 Extended Data Fig. 4: Genetic structure of European hunter-gatherers after the LGM. (a) Supervised ancestry 1123 modelling using non-negative least squares on IBD sharing profiles. Panels show estimated ancestry proportions for 1124 target individuals from genetic clusters representing European HGs, using different sets of increasingly proximal source 1125 groups. Individuals used as sources in a particular set are indicated with black crosses and coloured bars with 100% 1126 ancestry proportion. Black lines indicate 1 standard error for the respective ancestry component. (b) Residuals for 1127 model fit of target individuals from selected genetic clusters across different source sets. (c) Moon charts showing 1128 spatial distribution of ancestry proportions in European HGs deriving from four European source groups (set “hgEur2”; 1129 source origins shown with coloured symbol). Estimated ancestry proportions are indicated by both size and amount of 1130 fill of moon symbols. Note that ‘Italy_15000BP_9000 BP’ and ‘RussiaNW_11000BP_8000BP’ correspond to ‘WHG’ 1131 and ‘EHG’ labels used in previous studies. (d) Maps showing networks of highest between-cluster IBD sharing (top 10 1132 highest sharing per individual) for individuals from two genetic clusters representing Scandinavian HGs. See 1133 Supplementary Datas I and VII for details of individual sample IDs presented here. 1134 1135 Extended Data Fig. 5: Ancestry modelling for HG and Neolithic farmer-associated genetic clusters. Supervised 1136 ancestry modelling using non-negative least squares on IBD sharing profiles. Panels show estimated ancestry 1137 proportions of two global Eurasian clusters, corresponding to European HGs before 4,000 BP and individuals from 1138 Europe and Western Asia from around 10,000 BP until historical times, including Anatolian-associated (Neolithic) 1139 farmers, Caucasus HGs and recent individuals with genetic affinity to the Levant. Columns show results of modelling 1140 target individuals using three panels of increasingly distal source groups: “postBA”: Bronze Age and Neolithic source 1141 groups; “postNeol”, Bronze Age and later targets using Late Neolithic/early Bronze Age and earlier source groups; 1142 “deep”, Mesolithic and later targets using deep ancestry source groups. Individuals used as sources in a particular set 1143 are indicated with black crosses and coloured bars with 100% ancestry proportion. Black lines indicate 1 standard error 1144 for the respective ancestry component. 1145 1146 http://paperpile.com/b/xoiDNc/G0q5l http://paperpile.com/b/xoiDNc/G0q5l http://paperpile.com/b/xoiDNc/G0q5l http://paperpile.com/b/xoiDNc/G0q5l http://paperpile.com/b/xoiDNc/G0q5l 25 Extended Data Fig. 6: Ancestry modelling for post-Neolithic genetic clusters. Supervised ancestry modelling using 1147 non-negative least squares on IBD sharing profiles. Panels show estimated ancestry proportions of a global Eurasian 1148 cluster corresponding to European individuals after 5,000 BP, as well as pastoralist groups from the Eurasian steppe. 1149 Columns show results of modelling target individuals using three panels of increasingly distal source groups: “postBA”: 1150 Bronze Age and Neolithic source groups; “postNeol”, Bronze Age and later targets using Late Neolithic/early Bronze 1151 Age and earlier source groups; “deep”, Mesolithic and later targets using deep ancestry source groups. Individuals used 1152 as sources in a particular set are indicated with black crosses and coloured bars with 100% ancestry proportion. Black 1153 lines indicate 1 standard error for the respective ancestry component. 1154 1155 Extended Data Fig. 7: Ancestry modelling for genetic clusters east of the Urals. Supervised ancestry modelling 1156 using non-negative least squares on IBDaring profiles. Panels show estimated ancestry proportions of a global Eurasian 1157 cluster corresponding to Central, East and North Asian individuals with east Eurasian genetic affinities. Columns show 1158 results of modelling target individuals using three panels of increasingly distal source groups: “postBA”: Bronze Age 1159 and Neolithic source groups; “postNeol”, Bronze Age and later targets using Late Neolithic/early Bronze Age and 1160 earlier source groups; “deep”, Mesolithic and later targets using deep ancestry source groups. Individuals used as 1161 sources in a particular set are indicated with black crosses and coloured bars with 100% ancestry proportion. Black lines 1162 indicate 1 standard error for the respective ancestry component. 1163 1164 Extended Data Fig. 8: Dynamics of the Neolithic transition in Europe. (a) Supervised ancestry modelling using 1165 non-negative least squares on IBD sharing profiles. Panels show estimated ancestry proportions for target individuals 1166 from genetic clusters representing European Neolithic farmer individuals (“fEur” source set). Individuals used as 1167 sources in a particular set are indicated with black crosses and coloured bars with 100% ancestry proportion. Black lines 1168 indicate 1 standard error for the respective ancestry component. (b) Composition of HG ancestry proportions from 1169 different source groups in individuals with Neolithic farmer ancestry, shown as barplots. Grey bars represent 1170 contributions from a source with ancestry related to local HGs. (c) Moon charts showing spatial distribution of 1171 estimated ancestry proportions related to local HGs across Europe. Estimated ancestry proportions are indicated by size 1172 and amount of fill of moon symbols. Coloured areas indicate the geographic extent of individuals included as local 1173 sources in the respective regions. (d) Estimated time of admixture between local HG groups and Neolithic farmers. 1174 Black diamonds and error bars represent point estimate and standard errors of admixture time, coloured bars show 1175 temporal range of included target individuals. The time to admixture was adjusted backwards by the average age of 1176 individuals for each region. (e) Moon charts showing spatial distribution of estimated ancestry proportions derived from 1177 genetic clusters of early Neolithic European farmers (locations indicated with coloured symbols). Estimated ancestry 1178 proportions are indicated by size and amount of fill of moon symbols. Red symbols indicate individuals where standard 1179 errors exceed the point estimates for the respective ancestry source. 1180 1181 Extended Data Fig. 9: Dynamics of the steppe transition in Europe. (a) Estimated time of admixture between local 1182 hunter-gatherer groups and Neolithic farmers. Black diamonds and error bars represent point estimate and standard 1183 errors of admixture time, coloured bars show temporal range of included target individuals. The time to admixture was 1184 adjusted backwards by the average age of individuals for each region. (b) Moon charts showing spatial distribution of 1185 estimated ancestry proportions related to local Neolithic farmers across Europe. Estimated ancestry proportions are 1186 indicated by size and amount of fill of moon symbols. Coloured areas indicate the geographic extent of individuals 1187 included as local sources in the respective regions. (c) Maps showing networks of highest between-cluster IBD sharing 1188 (top 10 highest sharing per individual) for individuals from genetic cluster “Steppe_5000BP_4300BP” representing the 1189 major steppe ancestry source for Europeans. (d) Distributions of difference in estimated steppe-related ancestry 1190 proportions, using individuals from the genetic cluster “Steppe_5000BP_4300BP”, associated with either Yamnaya or 1191 Afansievo cultural contexts as separate sources. 1192 1193 Extended Data Fig. 10: Genetic transformations across the Eurasian steppe. (a) Moon charts showing spatial 1194 distribution of estimated ancestry proportions of Siberian HGs from the “deep” Siberian ancestry sources (names and 1195 locations indicated with coloured symbols). Estimated ancestry proportions are indicated by size and amount of fill of 1196 moon symbols. (b) Timelines of ancestry proportions from “postNeol” sources in Central and North Asian ancient 1197 individuals after 5,000 BP. Symbol shape and colour indicate the genetic cluster of each individual. Black lines indicate 1198 1 standard error. (c), (d). Difference in estimated steppe-related ancestry proportions, using individuals from genetic 1199 cluster “Steppe_5000BP_4300BP” associated with either Yamnaya or Afansievo cultural contexts as separate sources, 1200 as a function of time (c) or total estimated steppe-ancestry proportion (d). Individuals from genetic clusters of 1201 individuals associated with Okunevo (blue stars) or Sintashta/Andronovo (green diamonds) contexts are indicated with 1202 coloured symbols. 1203 1204 Extended Data Fig. 11: Patterns of co-ancestry. (a) Panels show within-cluster genetic relatedness over time, 1205 measured as the total length of genomic segments shared IBD between individuals. Results for both measures are 1206 26 shown separately for individuals from western versus eastern Eurasia. Small grey dots indicate estimates for individual 1207 pairs, with larger coloured symbols indicating median values within genetic clusters. Ranges of median values for major 1208 ancestry groups are indicated with labelled convex hulls. (b) Distribution of ROH lengths for 29 individuals with 1209 evidence for recent parental relatedness (>50 cM total in ROHs > 20 cM). (c) Karyogram showing genomic distribution 1210 of ROH in individual tem003, an ancient case of uniparental disomy for chromosome 2. Regions within ROH are 1211 indicated with blue colour. 1212 1213 Data availability 1214 All adapter-trimmed sequence data (fastq) for the samples sequenced in this study are publicly 1215 available on the European Nucleotide Archive under accession PRJEB64656, together with 1216 sequence alignment map files, aligned using human build GRCh37. The full analysis dataset 1217 including both imputed and pseudohaploid genotypes for all ancient individuals used in this study is 1218 available at https://doi.org/10.17894/ucph.d71a6a5a-8107-4fd9-9440-bdafdfe81455. Aggregated 1219 IBD-sharing data as well as hi-resolution versions of supplementary figures are available at Zenodo 1220 under accession 10.5281/zenodo.8196989. Previously published ancient genomic data used in this 1221 study is detailed in Supplementary Data VII, and are all already publicly available. 1222 Bioarchaeological data (including Accelerator Mass Spectrometry results) are included in the online 1223 supplementary materials of this submission. Map figures were created using Natural Earth Data (in 1224 Figs 1,2,3,6 and Extended Data Figs 1,3,4,8,9,10,11.). 1225 1226 Code availability 1227 All analyses relied upon available software which has been fully referenced in the manuscript and 1228 detailed in the relevant supplementary notes. A collection of R functions for IBD-based mixture 1229 model inference is available at https://github.com/martinsikora/mixmodel_ibd. 1230 1231 Acknowledgements 1232 We acknowledge Pia Bennike, involved in initiating this project, for her substantial contributions to its 1233 conception and to prehistoric research more broadly; she passed away in 2017. We thank Line Olsen and 1234 Pernille Selmer Olsen for administrative and technical assistance, respectively. We thank UK Biobank Ltd. 1235 for access to the UK Biobank genomic resource; thankfully acknowledge Illumina Inc. for collaboration. We 1236 thank Sturla Ellingvåg for assistance in sample access. E.W. thanks St. John’s College, Cambridge, for 1237 providing a stimulating environment of discussion and learning. The Lundbeck Foundation GeoGenetics 1238 Centre is supported by grants from the Lundbeck Foundation (R302-2018-2155, R155-2013-16338), the 1239 Novo Nordisk Foundation (NNF18SA0035006), the Wellcome Trust (214300), Carlsberg Foundation 1240 (CF18-0024), the Danish National Research Foundation (DNRF94, DNRF174), the University of 1241 Copenhagen (KU2016 programme), and Ferring Pharmaceuticals A/S to E.W.. This research has been 1242 conducted using the UK Biobank Resource and the iPSYCH Initiative, funded by the Lundbeck Foundation 1243 (R102-A9118 and R155-2014-1724). This work was further supported by the Swedish Foundation for 1244 Humanities and Social Sciences grant (Riksbankens Jubileumsfond M16-0455:1) to K.K. M.E.A. was 1245 supported by Marie Skłodowska-Curie Actions of the EU (grant no. 300554), The Villum Foundation (grant 1246 no. 10120) and Independent Research Fund Denmark (grant no. 7027-00147B). W.B. is supported by the 1247 Hanne and Torkel Weis-Fogh Fund (Department of Zoology, University of Cambridge); AP is funded by 1248 Wellcome grant WT214300, B.S.d.M and O.D. by the Swiss National Science Foundation (SFNS 1249 PP00P3_176977) and European Research Council (ERC 679330); R. Macleod by an SSHRC doctoral 1250 studentship grant (G101449: ‘Individual Life Histories in Long-Term Cultural Change’); G.R. by a Novo 1251 Nordisk Foundation Fellowship (gNNF20OC0062491); N.N.J. by Aarhus University Research Foundation; 1252 https://doi.org/10.17894/ucph.d71a6a5a-8107-4fd9-9440-bdafdfe81455 https://doi.org/10.5281/zenodo.8196989 https://github.com/martinsikora/mixmodel_ibd 27 B.S.P. by an ERC-Starter Grant 'NEOSEA' (grant no. 949424); H.S. by a Carlsberg Foundation Fellowship 1253 (CF19-0601); G.S. by Marie Skłodowska-Curie Individual Fellowship ‘PALAEO-ENEO’ (grant agreement 1254 number 751349); A.J. Schork by a Lundbeckfonden Fellowship (R335-2019-2318) and the National Institute 1255 on Aging (NIH award numbers U19AG023122, U24AG051129, and UH2AG064706); A.V.L. and I.V.S. by 1256 the Science Committee, Ministry of Education and Science of the Republic of Kazakhstan (AP08856317); 1257 B.G.R. and MGM by the Spanish Ministry of Science and Innovation (Project HAR2016-75605-R); C.M.-L. 1258 and O.R. by the Italian Ministry for the Universities (grants ‘2010-11 prot.2010EL8TXP_001 Biological and 1259 cultural heritage of the central-southern Italian population through 30 thousand Years’ and ‘2008 prot. 1260 2008B4J2HS_001 Origin and diffusion of farming in central-southern Italy: a molecular approach’); D.C.-S. 1261 and I.G.Z. by the Spanish Ministry of Science and Innovation (Project HAR2017-86262-P). D.C.S.G. 1262 acknowledges funding from the Generalitat Valenciana (CIDEGENT/2019/061) and the Spanish 1263 Government (EUR2020-112213); D.B. was supported by the NOMIS Foundation and Marie Skłodowska-1264 Curie Global Fellowship 'CUSP' (grant no. 846856); E.R.U. by the Science Committee, Ministry of 1265 Education and Science of the Republic of Kazakhstan (АР09261083: "Transcultural Communications in the 1266 Late Bronze Age (Western Siberia - Kazakhstan - Central Asia)"); E.C. by Villum Fonden (17649); J.E.A.T. 1267 by the Spanish Ministry of Economy and Competitiveness, (HAR2013‐46861‐R) and Generalitat Valenciana 1268 (Aico/ 2018/125 and Aico 2020/97); P.K. by the Russian Ministry of Science and Higher Education (Ural 1269 Federal University Program of Development within the Priority-2030 Program) and acknowledges the 1270 Museum of the Institute of Plant and Animal Ecology (UB RAS, Ekaterinburg). L.Y. acknowledges funding 1271 by the Science Committee of the Armenian Ministry of Education and Science (Project 21AG-1F025), L.O. 1272 by ERC Consolidator Grant ‘PEGASUS’ (agreement no. 681605); M. Sablin by the Russian Ministry of 1273 Science and Higher Education (075-15-2021-1069); N.C. by Historic Environment Scotland; S.V. and E.V. 1274 by the Russian Ministry of Science and Higher Education (075-15-2022-328); V.M. by the Science 1275 Committee, Ministry of Education and Science of the Republic of Kazakhstan (AR08856925). V.A. is 1276 supported by a Lundbeckfonden Fellowship (R335-2019-2318); P.H.S. by the National Institute of General 1277 Medical Sciences (R35GM142916); S.R. by the Novo Nordisk Foundation (NNF14CC0001); R.D. by the 1278 Wellcome Trust (WT214300); R.N. by the National Institute of General Medical Sciences (NIH grant 1279 R01GM138634); F. Racimo by a Villum Fonden Young Investigator Grant (no. 00025300). T.W. and V.A. 1280 are supported by the Lundbeck Foundation iPSYCH initiative (R248-2017-2003). 1281 1282 Author Information 1283 These authors contributed equally: Morten E. Allentoft, Martin Sikora, Alba Refoyo-Martínez, Evan K. 1284 Irving-Pease, Anders Fischer, William Barrie & Andrés Ingason 1285 1286 These authors equally supervised research: Morten E. Allentoft, Martin Sikora, Thorfinn Korneliussen, 1287 Richard Durbin, Rasmus Nielsen, Olivier Delaneau, Thomas Werge, Fernando Racimo, Kristian Kristiansen 1288 & Eske Willerslev 1289 1290 Contributions 1291 M.E.A., M.S., A.R.-M., E.K.I.-P., A.F., W.B., and A.I. contributed equally to this work. M.E.A., M.S., 1292 T.S.K., R.D., R.N., O.D., T.W., F. Racimo, K.K. and E.W. led the study. M.E.A., M.S., A.F., C.L.-F., R.N., 1293 T.W., K.K. and E.W. conceptualised the study. M.E.A., M.S., H.S., L.O., T.S.K., R.D., R.N., O.D., T.W., F. 1294 Racimo, K.K. and E.W. supervised the research. M.E.A., L.O., R.D., R.N., T.W., K.K. and E.W. acquired 1295 funding for research. A.F., J.S., K.G.S., M.L.S.J., M.U.H., A.A.T., A.C., A.Z., A.M.S., A.J.H, A.G., A.V.L., 1296 B.H.N., B.G.R, C.B., C.L., C.M-L., D.V., D.C.-S., D.L., D.N., D.C.S.-G., D.B., E.K., E.V.V., E.R.U., E. 1297 Kannegaard, F. Radina, H.D., I.G.Z., I.P., I.V.S., J.G., J.H., J.E.A.T., J.Z., J.V., K.B.P., K.T., L.N., L.L., 1298 28 L.M., L.Y., L.P., L. Sarti, L. Slimak, L.K., M.G.M., M. Silvestrini, M.V., M.S.N., M.P.R., M.H.S., M.P., 1299 M.C., M. Sablin, N.C., O.P., O.R., O.V.L., P.A., P.K., P.C., P. Ríos, P. Lotz, P. Lysdahl, P.P., P.B., P.d.B.D., 1300 P.V.P., P.P.M., P.W., R.V.S., R. Maring, R. Menduiña, R.B., R.T., S.V., S.W., S.B., S.N.S., S.A.S., S.H.A., 1301 T.D.P., T.J., Y.B.S., V.I.M., V.S., V.M, Y.M., I.M., O.G. and N.L. were involved in sample collection. 1302 M.E.A., M.S., A.R.-M., E.K.I.-P., W.B., A.I., J.S., A.P., B.S.d.M., M.I., L.V., A.J. Stern, C.G., F.E.Y, 1303 D.J.L., T.S.K., R.D., R.N., O.D., F. Racimo, K.K. and E.W. were involved in developing and applying 1304 methodology. M.E.A., J.S., C.G. and L.V. led the DNA laboratory work research component. K.G.S., A.F., 1305 M.E.A. led bioarchaeological data curation. M.E.A., M.S., A.R.-M., E.K.I.-P., W.B., A.I., A.P., B.S.d.M., 1306 B.S.P., A.S.H., R.A.H., T.V., H.M., A.M., A.V., A.B.N., P. Rasmussen, G.R., A. Ramsøe, A.S., A.J. Schork, 1307 A. Rosengren, C.J.M., I.A., L.Z., R.Maring, V.S., V.A., P.H.S, S.R., T.S.K., O.D. and F. Racimo undertook 1308 formal analyses of data. M.E.A., M.S., A.R.-M., E.K.I.-P., A.F., W.B., A.I., K.G.S., R. Macleod, D.J.L., 1309 P.H.S., T.S.K., F. Racimo and E.W. drafted the main text (M.E.A. and M.S. led this). M.E.A., M.S., A.R.-1310 M., E.K.I.-P., A.F., W.B., A.I., K.G.S., A.P., B.S.d.M., B.S.P, A.S.H., R. Macleod, R.A.H., T.V., M.F.M., 1311 A.B.N., M.U.H., P. Rasmussen, A.J. Stern, N.N.J., H.S., G.S., A. Ramsøe, A.S., A. Rosengren, A.K.O., 1312 A.B., A.C., A.G., A.V.L., A.B.G., C.J.M., D.C.S.-G., E. Kostyleva, E.R.U., E. Kannegaard, I.G.Z., I.P., 1313 I.V.S., J.G., J.H., J.E.A.T., L.Z, L.Y., L.P., L.K., M.B., M.G.M., M.V., M.P.R., M.J., N.B., O.V.L., O.C.U., 1314 P.K., P. Lysdahl, P.B., P.W., R.V.S., R. Maring, R.B., R.I., S.V., S.W., S.B., S.H.A., T.J., V.S., D.J.L., 1315 P.H.S., S.R., T.S.K., O.D. and F. Racimo drafted supplementary notes and materials. M.E.A., M.S., A.R.-M., 1316 E.K.I.-P., A.F., W.B., A.I., G.G.S., A.S.H., M.L.S.J., F.D., R. Macleod, L. Sørensen, P.O.N., R.A.H., T.V., 1317 H.M., A.M., N.N.J., H.S., A. Ramsøe, A.S., A.J. Schork, A. Ruter, A.K.O., B.H.N., B.G.R., D.C.-S., D.C.S.-1318 G., I.G.Z., I.P., J.G., J.E.A.T., L.Z., L.O., L.K., M.G.M., P.d.B.D., R.I., S.A.S., D.J.L., I.M., O.G., P.H.S., 1319 T.S.K., R.D., R.N., O.D., T.W., F. Racimo, K.K. and E.W. were involved in reviewing drafts and editing 1320 (M.E.A., M.S., A.F., K.G.S., R. Macleod, and E.W. led this, and subsequent finalisation of the study). All 1321 co-authors read, commented on, and agreed upon the submitted manuscript. 1322 1323 Corresponding authors 1324 1325 Correspondence to Morten E. Allentoft (morten.allentoft@curtin.edu.au), Martin Sikora 1326 (martin.sikora@sund.ku.dk), Eske Willerslev (ew482@cam.ac.uk). 1327 Ethics declarations 1328 1329 Competing interests 1330 1331 The authors declare no competing interests. 1332 mailto:morten.allentoft@curtin.edu.au mailto:martin.sikora@sund.ku.dk mailto:ew482@cam.ac.uk Age (BP) a b Southern Europe Western Europe Northern Europe Central / Eastern Europe Western Asia Central Asia North Asia −25000 −20000 −15000 −10000 −5000 0 −0.02 0.00 0.02 0.04 0.06 −0.01 0.00 0.01 0.02 0.03 PC1 (5.38%) P C 2 ( 1 .3 1 % ) −0.06 −0.03 0.00 0.03 0.06 −0.06 −0.03 0.00 0.03 PC1 (1.37%) P C 2 ( 0 .6 5 % ) Americas Western Asia E u ro p e Neolithic Steppe cline E u ro p e a n h u n te r- g a th e re r c lin e East West North Asia BA Steppe cline c d 5000 10000 15000 Average age region Southern Europe Western Europe Northern Europe Central/Eastern Europe Western Asia Central Asia South Asia Southeast Asia East Asia North Asia North America South America Australasia Melanesia European farmer cline Early Late Iran / Caucasus Steppe Levant Europe post-Neolithic Europe Southeast Italy Iberia France Britain / Ireland Europe Central / East Denmark Sweden / Norway Baltic Europe East 12,000 11,000 10,000 9,000 8,000 7,000 6,000 5,000 4,000 3,000 2,000 1,000 0 Years before present A n c e s tr y p ro p o rt io n a b HG Siberia Forest Steppe HG Siberia NE HG North Asia HG Upper Paleolithic HG Europe S HG Russia HG Ukraine HG Middle Don Farmer Anatolia HG Caucasus 0 10 20 30 40 0 25 50 75 100 Steppe−related ancestry (%) G A C F a rm e r− re la te d a n c e s tr y ( % ) −20 −10 0 10 20 −5,000 −4,000 −3,000 −2,000 −1,000 0 Time (years before present) S te p p e -r e la te d a n c e s tr y d if fe re n c e ( % ) Early pulse of G AC -related adm ixture dilutes Steppe ancestry in N E Europe Dil uti on of bo th Ste pp e a nd G AC an ce str y thr ou gh su bs eq ue nt ad mi xtu re pro ce ss se s A fa n a s ie v o Y a m n a y a a b EuropeNE_4800BP_3000BP Europe_4500BP_2000BP Poland_4400BP Scandinavia_4200BP_3200BP Scandinavia_4600BP_3800BP EuropeE_4000BP_2500BP EuropeNW_4000BP_500BP Scandinavia_4000BP_3000BP Estonia_3000BP_2500BP H G n o n − lo c a l E n o n − lo c a l W −5000 −4000 −3000 −2000 −1000 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 Age (Years BP) A n c e s tr y ( % ) MiddleDon_7500BP Turkmenistan_7000BP_5000BP LevantEuropeS_4700BP_1700BP Poland_5000BP_4700BP Steppe_5000BP_4300BP Thailand_1700BP SteppeC_8300BP_7000BP Botai_5600BP_5100BP SteppeC_6700BP_4600BP SteppeCE_7000BP_3600BP SiberiaNE_9800BP Amur_7500BP Baikal_8000BP_7200BP before 9,000 cal. BP 9,000 - 7,000 cal. BP 7,000 - 5,000 cal. BP 5,000 - 3,000 cal. BP Article File Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6