|
Post by Admin on May 29, 2020 2:53:26 GMT
The Near East was a crossroad for the ancient world's greatest civilizations, and invasions over centuries caused enormous changes in cultures, religions and languages. However, a new study of the DNA of ancient skeletons spanning 4,000 years has revealed that most of these changes had no lasting effect on the genetics of the local population of Beirut. Whilst the invasions and conquests may have been revolutionary for the elite rulers, researchers at the Wellcome Sanger Institute, University of Birmingham, French Institute of the Near East in Lebanon and their collaborators found only three time periods that had any impact on the long-term genetics of the ordinary people. These were the beginning of the Iron Age, the arrival of Alexander the Great, and the domination of the Ottoman Empire. Reported today in the American Journal of Human Genetics, the study shows the value of using genetics alongside archaeology to help understand what could be happening in the lives of ordinary people throughout history. Over the centuries, the Levant has had many different rulers, including the Egyptians, Babylonians, Assyrians, Persians, Greeks, Romans, Crusaders, Arabs, and Ottomans. Most of these had permanent cultural effects on the local population, including changes to religion and even languages, as shown by the historical records and archaeological findings. However, despite this, previous research showed that present-day local people in Lebanon were mainly descended from local people in the Bronze Age (2100-1500 BCE), with 90 per cent of their genetic make-up coming from around 4,000 years ago, and very few lasting traces of even the Crusaders invasion around the 11th-13th Century. To understand this potential contradiction and build a picture of the genetic history of ordinary people in the region, the researchers studied the DNA of ancient skeletons through 4,000 years. The team sequenced the genomes of 19 ancient people who lived in Lebanon between 800BCE and 200CE, and by combining with previous ancient and modern data, created an 8-point time line across the millennia. Scientists detected lasting genetic changes in the local people from just three time periods—during the beginning of the Iron Age (about 1,000 BCE), the arrival of Alexander the Great (beginning 330 BCE), and the domination of the Ottoman Empire (1516 CE) - but not from the other times. Dr. Marc Haber, first author from the University of Birmingham and previously from the Wellcome Sanger Institute, said: "We revealed a genetic history of the area across 4,000 years, with a time-point approximately every 500 years. This showed us that despite the huge cultural changes that were occurring during this period, there were only a few times that the genetics of the general population changed enough to affect the ordinary people." The study revealed that some people did mix and form families with people from other cultures. One burial site was found to contain the remains of an Egyptian mother, and her son whose father had Egyptian and Lebanese ancestry. However, this cosmopolitan mixing did not seem to be widespread. Historical evidence is based on archaeological findings and written records, but these are biased towards the elite rulers and people with money and influence, as they have far more resources and write the history. It can be difficult to understand the lives of the ordinary people. Dr. Joyce Nassar, an author on the paper and archaeologist from the French Institute of the Near East, Lebanon, said: "This study is really exciting, as the genetic evidence is helping us to interpret what we find. Some people might think that when a land was invaded, that the population would change. But this study shows it isn't that simple, and reveals there was only limited biological mixing, despite the cultural and political influence of the invasions." The skeletons came from four archaeological excavation sites in Beirut, which were discovered during building projects in the Lebanese capital city and rescued by the Directorate General of Antiquities. The archaeologists and researchers then worked together to transfer the bones to a laboratory in Estonia dedicated to ancient DNA, where the surviving ancient DNA was extracted from the temporal bone in the skulls. The DNA was then sequenced and analysed at the Sanger Institute. Recent advances in DNA extraction and sequencing technology made studying the ancient and damaged DNA possible. Dr. Chris Tyler Smith, senior author on the paper and previously from the Wellcome Sanger Institute, said: "We see that people like the Egyptians and the Crusaders came to Lebanon, lived, raised families and died there. Their DNA sequences reveal this, but a little while later, there may be no trace of their genetics in the local population. Our study shows the power of ancient DNA to give new information about the human past, that complements the available historical records, and reveals the benefits of archaeologists and geneticists working together to understand historical events."
|
|
|
Post by Admin on May 29, 2020 4:38:28 GMT
The researchers, made up of an international team of scientists including Harvard anthropology professor Christina Warinner, looked at DNA data from 110 skeletal remains in West Asia dated 3,000 to 7,500 years ago. The remains came from archaeological sites in the Anatolia (present-day Turkey), the Northern Levant which includes countries on the Mediterranean coast such as Israel and Jordan, and countries in the Southern Caucasus which include present-day Armenia and Azerbaijan. Based on their analysis, the scientists describe two genomic events that occurred around 8,500 years ago and 4,000 years ago that pointed to long-term genetic mixing in the region and subtle population movements within the area, shedding light on a long-standing question. “Within this geographic scope, you have a number of distinct populations, distinct ideological groups that are interacting quite a lot and it hasn’t really been clear to what degree people are actually moving or if this is simply just a high contact area from trade,” said Warinner, assistant professor of anthropology at the Faculty of Arts and Sciences and the Sally Starling Seaver Assistant Professor at the Radcliffe Institute for Advanced Study. “What we can see is that rather than this period being characterized by dramatic migrations or conquest, what we see is the slow mixing of different populations, the slow mixing of ideas, and it’s percolating out of this melting pot that we see the rise of urbanism — the rise of cities,” The study was led by the Max Planck-Harvard Research center for the Archaeoscience of the Ancient Mediterranean and published in the journal Cell. Warinner was a senior author on the paper. Historically, Western Asia, which includes the modern-day Middle East, is one of the world’s most important geographical locations. Early on it not only created some of humanity’s earliest cities but its early trade routes laid the foundation for what would become the Silk Road, a route that commercially linked Asia, Africa, and Europe. Even prior to being connected with other regions, however, populations across Western Asia had already developed their own distinct traditions and systems of social organization and complexity. The areas studied in this paper played major roles in this development from early farming to pastoral communities to early state-level societies. With the study, the researchers wanted to fill in some of the anthropological gaps between the origins of agriculture and of cities to better understand these different communities came together to eventually form cities. “What we see in archeology is that the interconnectivity within Western Asia increased and areas such as Anatolia, the Northern Levant, and the Caucasus became a hub for [the] exchange of ideas and material culture,” said Eirini Skourtanioti, a Ph.D. student at the Max Plank Institute and the lead author of the study, in a video accompanying the release of the paper. “The goal of our study was to understand the role of human mobility throughout this process.” A partial map of West Asia, which includes Anatolia (present-day Turkey), the Northern Levant, and the Southern Caucasus. An international team of researchers showed populations from Anatolia and the Caucasus started genetically mixing around 6,500 BC and that small migration events from Mesopotamia 4,000 years ago brought further genetic mixture to the region. The orange marker shows the route from Central Asia. DNA from a lone ancient woman revealed proof of long distance migration during the late Bronze age about 4,000 years ago from Central Asia to the Mediterranean Coast. Image Credit : Max Planck-Harvard Research center for the Archaeoscience of the Ancient Mediterranean The researchers included an international team of authors from many disciplines and countries, including Australia, Azerbaijan, France, Italy, Germany, South Korea, Turkey, and the United States. They gathered the 110 ancient remains and took samples from their teeth and part of the temporal bone called the petrous, which is part of the inner ear. The samples from the skeletons were all previously excavated and were housed in different museums and labs around the world. The genetic analysis was all conducted by scientists at the Max Planck Institute, including Warinner. In the paper, the authors outline how approximately 8,500 years ago, populations across Anatolia and the Southern Caucasus began genetically mixing. It resulted in a gradual change in genetic profile that over a thousand years slowly spread across the both areas and entered into what is now Northern Iraq. Known as a cline in genetics, this mixture indicated to the researchers ongoing human mobility in the area and the development of a regional genetic melting pot in Anatolia and its surrounding areas. The other shift researchers detected wasn’t as gradual. They looked at samples from the ancient cities of Alalakh and Ebla in what is today southern Turkey and northern Syria and saw that around 4,000 years ago the Northern Levant experienced a relatively sudden introduction of new people. The subtle genetic shifts points to a mass migration event. The timing of this migration corresponds with a massive drought in Northern Mesopotamia. It is likely where the migrants that entered the Northern Levant area originated from. The scientists can’t be sure because there are currently no well preserved genomes for Mesopotamia. Along with findings on interconnectivity in the region, the paper presents new information about long distance migration during the late Bronze age about 4,000 years ago. Researchers determined that a lone corpse, found buried in a well, genetically belonged in Central Asia at the time, not at a site that is part of present-day Turkey. “We can’t exactly know her story, but we can piece together a lot of information that suggests that either she or her ancestors were fairly recent migrants from Central Asia,” said Warinner, who is also a group leader in the Department of Archaeogenetics at the Max Planck Institute. “We don’t know the context in which they arrived in the Eastern Mediterranean but this is a period of increasing connectivity in this part of the world.” The corpse had many injuries and the way she was buried indicates a violent death. Warinner hopes more genomic analysis can play some type of role in unraveling the ancient woman’s story. For Warinner, who earned her master’s in 2008 and her Ph.D. in 2010 from the Graduate School of Arts and Sciences, these types of studies are proof of the insights DNA analysis can provide when more traditional clues don’t tell the full story. “What’s really interesting is that we see these populations are mixing genetically long before we see clear material culture evidence of this — so, long before we see direct evidence in pottery or tools or any of these more conventional archaeological evidence artifacts,” Warinner said. “That’s important because sometimes we’re limited in how we see the past. We see the past through artifacts, through the evidence people leave behind. But sometimes events are happening that don’t leave traces in conventional ways, so by using genetics, we were able to access this much earlier mixing of populations that wasn’t apparent before.” Genomic History of Neolithic to Bronze Age Anatolia, Northern Levant, and Southern Caucasus doi.org/10.1016/j.cell.2020.04.044Summary Here, we report genome-wide data analyses from 110 ancient Near Eastern individuals spanning the Late Neolithic to Late Bronze Age, a period characterized by intense interregional interactions for the Near East. We find that 6th millennium BCE populations of North/Central Anatolia and the Southern Caucasus shared mixed ancestry on a genetic cline that formed during the Neolithic between Western Anatolia and regions in today’s Southern Caucasus/Zagros. During the Late Chalcolithic and/or the Early Bronze Age, more than half of the Northern Levantine gene pool was replaced, while in the rest of Anatolia and the Southern Caucasus, we document genetic continuity with only transient gene flow. Additionally, we reveal a genetically distinct individual within the Late Bronze Age Northern Levant. Overall, our study uncovers multiple scales of population dynamics through time, from extensive admixture during the Neolithic period to long-distance mobility within the globalized societies of the Late Bronze Age.
|
|
|
Post by Admin on May 29, 2020 21:20:07 GMT
A Genetic History of the Near East from an aDNA Time Course Sampling Eight Points in the Past 4,000 Years
The Iron and Classical Ages in the Near East were marked by population expansions carrying cultural transformations that shaped human history, but the genetic impact of these events on the people who lived through them is little-known. Here, we sequenced the whole genomes of 19 individuals who each lived during one of four time periods between 800 BCE and 200 CE in Beirut on the Eastern Mediterranean coast at the center of the ancient world’s great civilizations. We combined these data with published data to traverse eight archaeological periods and observed any genetic changes as they arose. During the Iron Age (∼1000 BCE), people with Anatolian and South-East European ancestry admixed with people in the Near East. The region was then conquered by the Persians (539 BCE), who facilitated movement exemplified in Beirut by an ancient family with Egyptian-Lebanese admixed members. But the genetic impact at a population level does not appear until the time of Alexander the Great (beginning 330 BCE), when a fusion of Asian and Near Easterner ancestry can be seen, paralleling the cultural fusion that appears in the archaeological records from this period. The Romans then conquered the region (31 BCE) but had little genetic impact over their 600 years of rule. Finally, during the Ottoman rule (beginning 1516 CE), Caucasus-related ancestry penetrated the Near East. Thus, in the past 4,000 years, three limited admixture events detectably impacted the population, complementing the historical records of this culturally complex region dominated by the elite with genetic insights from the general population.
The ancient Near East has been at the center of interaction between the ancient world’s civilizations and was ruled at different times by the Egyptians, Hittites, Assyrians, Babylonians, Persians, Greeks, Romans, Arabs, Crusaders, Mamluks, and Ottomans, most of whom left a permanent cultural impact on the local population. However, their genetic contribution is not as evident: our previous ancient DNA (aDNA) work showed that people who live in the Near East today derive ∼90% of their ancestry from the local Bronze Age population that preceded all of the aforementioned historical conquests.1 These results might appear to challenge the historical records of population movements, colonization, and admixture with the locals throughout history. For example, in 1307 CE, the Mamluks divided Lebanon’s coast among 300 newly introduced Turkoman families, and a few centuries earlier the Romans had declared Beirut and Baalbek in Lebanon as colonies and garrison towns;2 additionally the names of Hellenistic army soldiers and their descendants in Lebanon can still be read today from inscriptions on funerary stela found in Sidon.3 Similarly, our analysis of aDNA from a Crusader burial site in Lebanon showed that immigration to the Near East and admixture with the locals was common, and for a period, a heterogeneous population of Europeans, locals, and their admixed descendants lived in the Near East.4 However, this admixture appears to not have been widespread enough to leave a permanent genetic impact on the local population, and subsequent mixing with people carrying the local ancestry “diluted” the ancestry of the Crusaders in Near Eastern genomes to undetectable levels. The example of the Crusaders might illustrate why, even after numerous conquests and immigrations, the Near Eastern Bronze Age ancestry still dominates present-day Near Eastern genomes. Thus, two outstanding questions emerge from the previous aDNA studies: (1) were transient admixture events a common occurrence in the history of the Near East, or was the Crusaders period an exception, and (2) because present-day Near Easterners derive most but not all of their ancestry from the local Bronze Age population, which post-Bronze Age events contributed to the genetic diversity we observe today in the Near East.
To address these questions, we have now sequenced the genomes of ancient individuals who lived between 800 BCE and 200 CE at one of four different time periods: the Iron Age II (1000–539 BCE), the Iron Age III (539–330 BCE), the Hellenistic period (330–31 BCE), and the early Roman period (31 BCE–200 CE) (Table 1). These data, together with previous data we generated from individuals from the same region from the Middle Bronze Age (2100–1550 BCE), the late Roman period (200–634 CE), the Crusader period (1099–1291 CE), and the present-day provide a genetic representation of the Near East in a time series spanning the past 4,000 years (Table S1).
Table 1. Samples Analyzed in this Study
ENA Number ID Excavation Site Period Date (Calibrated) Mapped Read % Coverage Genomic SNPs Overlapping with Published aDNA Data ERS4542976 SFI-56 Beirut SFI-415 Iron Age II – 15 0.7 568,628 ERS4542991 SFI-55 Beirut SFI-415 Iron Age II – 8 0.4 408,015 ERS4542962 SFI-43 Beirut SFI-1075 Iron Age III 567 BCE–404 BCE 17 0.5 428,888 ERS4542967 SFI-50 Beirut SFI-1075 Iron Age III – 31 1 703,041 ERS4542969 SFI-36 Beirut SFI-1075 Iron Age III – 19 0.8 590,514 ERS4542989 SFI-42 Beirut SFI-1075 Iron Age III 540 BCE–396 BCE 13 0.5 440,585 ERS4542964 SFI-45 Beirut SFI-1075 Iron Age III – 24 0.6 478,277 ERS4542984 SFI-34 Beirut SFI-1075 Iron Age III – 27 1.7 933,032 ERS4542983 SFI-35 Beirut SFI-1075 Iron Age III – 5 0.3 321,527 ERS4542988 SFI-39 Beirut SFI-1075 Iron Age III – 13 0.7 567,178 ERS4542990 SFI-44 Beirut SFI-1075 Iron Age III – 41 1.6 889,705 ERS4542987 SFI-47 Beirut SFI-1075 Iron Age III – 23 1.1 747,390 ERS4542979 SFI-20 Beirut SFI-477 Hellenistic 199 BCE–37 BCE 13 0.8 691,379 ERS4542972 SFI-5 Beirut SFI-477 Hellenistic 234 BCE–92 BCE 3 0.1 140,660 ERS4542974 SFI-12 Beirut SFI-477 Hellenistic 209 BCE–89 BCE 2 0.1 106,051 ERS4542980 SFI-24 Beirut SFI-1106 early Roman 55 BCE–58 CE 39 3.3 1,093,459 ERS4542982 SFI-33 Beirut SFI-1106 early Roman 48 CE–222 CE 43 3.3 1,087,690 ERS4542973 SFI-11 Beirut SFI-477 early Roman 119 BCE–27 CE 2 0.1 132,450 ERS4542977 SFI-15 Beirut SFI-477 early Roman 176 BCE–3 CE 28 1.4 915,901
We sampled the petrous portion of the temporal bones from 67 individuals buried in Beirut (Figures S1 and S2), a city on the Eastern Mediterranean coast that has had continuous settlement dating back 5,000 years and that is the capital of modern-day Lebanon. We extracted DNA and built double-stranded libraries according to published protocols5, 6, 7 and sequenced the libraries on Illumina HiSeq 2500 and HiSeq 4000 platforms with 2 × 75 bp reads. We processed the sequences by using PALEOMIX8 as described previously4 and mapped the merged sequences to the hs37d5 reference sequence (see Supplemental Methods). We found 19 samples that had 2%–43% endogenous DNA with post-mortem damage patterns typical of ancient DNA (Figure S3), and subsequent sequencing of these libraries resulted in genomic coverage between 0.1× and 3.3× (Table 1). We estimated contamination from the X chromosomes of males and the mtDNA genome of all individuals9,10 and found that the sequence data were minimally contaminated (Tables S2 and S3).
We combined the new data with published ancient and modern data, creating two datasets: set 1 included 2,012 modern humans1,11, 12, 13, 14 and 914 ancient individuals6,15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37 with 815,791 SNPs, and set 2 consisted of 2,788 modern humans24,38,39 and 914 ancient individuals with 539,766 SNPs (see Supplemental Methods).
|
|
|
Post by Admin on May 30, 2020 1:16:03 GMT
We then estimated kinship40 among our samples and found individuals SFI-43 (female) and SFI-44 (male), who lived around 500 BCE during the Iron Age III under the Persian rule, were first-degree relatives (Figure S4) and shared the same mtDNA haplogroup, T2C1 (Table S4). We kept these two individuals in the dataset for the following test and projected all ancient samples in set 2 onto a principal component analysis (PCA)41 plot based on variation in modern West, Central and South Eurasians (Figures 1 and S5). The plot differentiates between populations from the Near East, Europe, Caucasus, Russian Steppe, Central and South Asia. The ancient Lebanese (i.e., ancient individuals who lived in what is today known as Lebanon) clustered with the modern and ancient Near Easterners: the new samples clustered between the Bronze Age population (Sidon_BA) and modern Lebanese. The two first-degree relatives, SFI-43 and SFI-44, appeared as outliers and did not cluster with their contemporaries, but instead were positioned close to the Bronze Age samples. We wanted to test whether these two individuals had a genetic affinity to a population other than the ancient Lebanese. Thus, using qpWave,42,43 we selected 11 outgroups (see Supplemental Methods) that have different relationships with the populations found in set 1 and tested whether SFI-43 and SFI-44 formed a clade with any of the populations (including the ancient Lebanese) in our dataset. We found that SFI-43 only formed a clade with ancient Egyptians (Table S5), implying that she shared all of her ancestry with them or a genetically equivalent population. On the other hand, SFI-44’s ancestry appeared to be more complex because he did not form a clade with any population in our dataset, yet he appeared to share ancestry with SFI-43, ancient Egyptians, and ancient Levantines (Table S5). To better understand the relationship of SFI-43 and SFI-44 with the Lebanese and Egyptians, we projected the ancient Lebanese and ancient Egyptians onto a PCA constructed with the variation found in their modern populations. SFI-43 and SFI-44 clustered with the ancient Egyptians and were positioned between modern or ancient Lebanese and modern Egyptians, but SFI-44 was positioned closer than SFI-43 to the Lebanese (Figure S6). Because SFI-43 and SFI-44 are first-degree relatives but appear to have differences in their genetic ancestry, we tested whether SFI-44 can be modeled as a mixture of ancestries deriving from SFI-43 and any other individuals or populations in our dataset by using qpAdm.42 We found that SFI-43 could be modeled as deriving ∼70% of his ancestry from a population related to SFI-44 and ∼30% from a population related to ancient Levantines (Table S6). But these ancestry proportions do not reflect the first-degree relationship that the two individuals shared unless more than one mixture event had occurred in the family, so we created a simulated hybrid genome that represents a first-generation mixture between an ancient Egyptian and an ancient Lebanese and tested whether SFI-44 could be modeled as descending from a mixture between SFI-43 and the hybrid genome. The model showed that SFI-44 derived ∼50% of his ancestry from SFI-43 and ∼50% from an individual whose ancestry was similar to that of the hybrid genome (Table S6). Thus, these results suggest that SFI-43 was an Egyptian woman and SFI-44 was her son from a man who himself had both Egyptian and Lebanese ancestries. The structure of this family in Lebanon highlights population movements and the heterogeneous society that existed at that time, but additional sampling is needed if we are to understand whether this cross-cultural mixing was common or whether our samples were exceptional. We removed SFI-43 and SFI-44 from all following analyses in which local individuals were grouped to represent their respective time periods. Figure 1. Principal Components Analysis of West, Central, and South Eurasians Eigenvectors were inferred with present-day populations (light-colored points in the background of the plot), and the ancient samples (colored solid shapes in the foreground of the plot) were projected onto the plot. Having genetic representation from eight consecutive time periods (Figure 2A), we were able to test whether two populations that were successive in time formed a clade and derived all of their ancestry from a shared ancestral population or whether subsequent admixture had occurred and the two populations consequently lost their clade relationship. We started by computing f4-statistics of the form f4(Lebanon Period1, Lebanon Period2; Ancient, Chimpanzee), in which a result significantly different from zero could indicate that genetic changes related to “Ancient” (an ancient population in our dataset) have occurred between two successive periods in Lebanon. We found that significant genetic changes that were marked by an increase in Eurasian ancestry related to ancient Europeans and ancient Central Asians occurred after the Bronze Age and starting from the Iron Age II (Figure S7A). We did not observe significant genetic differences between the Iron Age II and Iron Age III populations in this test (Figure S7B), and thus, we merged our samples from these two periods into one population (Figure S7C) and used qpAdm (see Supplemental Methods) to explore possible Iron Age admixture models (Tables 2 and S7). We found that the Lebanese Iron Age population can be modeled as a mixture of the local Bronze Age population (63%–88%) and a population related to ancient Anatolians or ancient South-Eastern Europeans (12%–37%) (Table 2 and Figure 2B). We replicated these results by running DyStruct44 with 166,693 transversions present in set 1 and showed that a Steppe-like ancestry, typically found in Europeans, appears in the Near East starting from the Iron Age II (Figure 2D). A potential source of this exogenous ancestry could be the Sea Peoples, a seafaring group of people with a disputed origin who attacked the Eastern Mediterranean and Egypt after the Bronze Age (1200–900 BCE). One of our successful models for admixture involved an ancestry source related to the Ashkelon (a city situated ∼170 miles south of the Beirut sites) Iron Age I population, which was previously identified as possibly descending from Sea-Peoples-related admixture.18 In addition, according to ancient Egyptian texts and archaeology, the Sea Peoples conquered the Levant but failed to conquer the Egyptians. Therefore, we tested whether the Eurasian gene flow to Lebanon during the Iron Age had also reached ancient Egypt by quantifying the Steppe ancestry in both regions at that time and found f4(Sidon_BA, Beirut_IAII; Steppe_EMBA, Chimp) is significantly negative (Z score = −4.13), but f4(Sidon_BA, Egypt_prePtolemaic; Steppe_EMBA, Chimp) has a value not significantly different from zero (Z score = 0.317), suggesting that either ancient Egypt did not receive the Eurasian gene flow that the Levant received during the Iron Age or that the Eurasian ancestry was replaced in Egypt as in Ashkelon, where in contrast to the Beirut_IAII, the European-related ancestry was no longer significant in the Ashkelon Iron Age II population.18 Additional Iron Age samples from the Levant coast and Egypt could reveal whether the Iron Age admixture had a north to south cline as a result of the location of the source populations or from differences in the scale of the successful migrations to the north or south of the Levant during this period. Figure 2. Admixture in Ancient Lebanon (A) Historical context of the studied samples. Horizontal lines indicate the time period of a sampled population, and the blue lozenges represent newly sequenced samples. (B and C) Locations of the source populations we used in qpAdm to test for admixture at the Iron Age (B) and the Hellenistic/early Roman period (C). The black lozenge on each map shows Lebanon’s location. Points represent modern populations in the dataset, whereas triangles represent ancient populations. Increased intensity of the red color indicates a higher p value for the model involving the source population (this should not be interpreted as an indication of the best model). We set the p values of the models that can be rejected to zero. (D) A DyStruct run with 166,693 transversions found in set 1 across nine time points. We show the plot of K = 6, which reveals an ancestral component (red) related to the Bronze Age Steppe population appearing in the Near East after the Bronze Age. (E) Haplotype segments shared between the ancient Lebanese and global modern populations. The heatmap is based on ChromoPainter’s co-ancestry matrix, and we averaged values from the modern populations over all individuals in the population. We scaled the heatmap by row to highlight the differences between the ancient individuals. Two Hellenistic individuals and one early Roman individual showed excess haplotype sharing with Central and South Asian populations compared with that of other ancient Lebanese individuals, whereas individuals SFI-43 and SFI-44 shared more segments with Africans and Egyptians. We counted between 19,073 (blue) and 19,659 (red) shared haplotype chunks in the dataset. Table 2. Modeling Populations from the Iron Age and Antiquity as a Mixture of the Preceding Population, A, and Any Global Ancient Population, B Test A B p Value for Rank = 1 A B Std. Error Mixture Proportions Beirut_IA Sidon_BA Anatolia_MLBA 4.44 × 10−01 0.63 0.37 0.06 Beirut_IA Sidon_BA Ashkelon_IAI 4.29 × 10−01 0.69 0.31 0.05 Beirut_IA Sidon_BA Anatolia_EBA 3.38 × 10−01 0.80 0.20 0.03 Beirut_IA Sidon_BA Mycenaean 2.17 × 10−01 0.77 0.23 0.04 Beirut_IA Sidon_BA Minoan_Odigitria 1.32 × 10−01 0.80 0.20 0.04 Beirut_HER Beirut_IA Butkara_H 4.93 × 10−01 0.92 0.08 0.01 Beirut_HER Beirut_IA Aligrama2_IA 4.46 × 10−01 0.93 0.07 0.01 Beirut_HER Beirut_IA Indus_Periphery 3.88 × 10−01 0.93 0.07 0.01 Beirut_HER Beirut_IA Swat_H 3.24 × 10−01 0.92 0.08 0.01 Beirut_HER Beirut_IA SPGT_IA 2.65 × 10−01 0.93 0.07 0.01 We show the top five models for each test based on their p value for the rank = 1 matrix. A p value > 0.05 indicates the model cannot be rejected. We removed infeasible models with negative proportions from the table. Beirut_IA included individuals from the Iron Age II and Iron Age III periods and can be modeled as a mixture of the local Bronze Age population and a population related to ancient Anatolians or ancient South-Eastern Europeans. Beirut_HER included individuals from the Hellenistic and early Roman periods and can be modeled as a mixture of the local population Beirut_IA and an ancient Central and South Asian population. The second genetic change in ancient Lebanon can be observed during the Hellenistic and early Roman periods. We merged individuals from these two periods into one population (Beirut_HER) because several individuals had overlapping radiocarbon dates and the f4-statistics showed symmetry between the Beirut_Hellenistic and Beirut_ERoman populations (Figure S8). We found that the Hellenistic and early Roman population can be modeled as a mixture of the local population, Beirut_IA (88%–94%), and a Central/South Asian population (6%–12%) (Tables 2 and S8 and Figure 2C). We then analyzed haplotype segments shared between the ancient Lebanese and modern populations in set 2 by using ChromoPainter44 on 2.5 million imputed SNPs and found that two Hellenistic individuals (SFI-5 and SFI-12) and one early Roman individual (SFI-11) had excess haplotype sharing with Central and South Asians (Figures 2E and S9), thus confirming the qpAdm results. The relationship of ancient Lebanon with Central and South Asia also manifests in the presence of haplogroup L1a1-M27 among the modern Lebanese Y chromosome lineages (Figure S10). Haplogroup L1a1-M27 is common today in Central and South Asia but rare elsewhere (in the 1000 Genomes Project,45 this lineage was found exclusively in Sri Lankan Tamil from the UK [STU], Punjabi from Lahore, Pakistan [PJL], Indian Telugu from the UK [ITU], Gujarati Indian from Houston, Texas [GIH], and Bengali from Bangladesh [BEB]). We tested46 (see Supplemental Methods) the coalescence of the five L1a1-M27 Lebanese chromosomes and found that they all derived from a man who lived around 450 BCE–50 CE, a time interval overlapping with the Hellenistic period (Figure S10). The presence of the Central/South Asian ancestry in Lebanon during the Hellenistic period mirrors the connected geography under the rule of Alexander the Great’s empire, which had also assimilated the Achaemenid Empire that preceded it and thus maintained a connection between the West and East for five centuries. These large contiguous empires thus facilitated the movement and mixture of people as seen directly by the Egyptian-Lebanese family and the admixed individuals reported here who lived in the Near East at that time.
|
|
|
Post by Admin on May 30, 2020 19:25:39 GMT
We next tested the genetic changes between the Hellenistic/early Roman period and the late Roman period (Qed_LRoman) and found little genetic differences from the f4-statistics (Figure S11), which is notable because during this period there was significant population movement between the Near East and Europe, as identified from the genomes of ancient Near Easterners found in Rome at that time.16 When we model Qed_LRoman as a mixture of the Hellenestic/early Roman period population and another ancient population, we find successful models involving ancient Anatolians and South-Eastern Europeans (Table S9). However, because this ancestry was already present in Lebanon starting from the Iron Age, its excess in Qed_LRoman could be from population structure, especially because the Qed_LRoman samples were from a remote mountainous region, whereas the Hellenistic/early Roman samples were from the coast, and in addition, we found that the admixture models were not significant when Beirut_IA was used as the source of the local ancestry, showing that Qed_LRoman derived all of its ancestry from preceding local populations (Table S9). From the late Roman period to the medieval period, we detect an increase in African ancestry (Figure S11B), but that increase remains slightly below statistical significance (Z score = −2.4) and accounts for ∼2.9% of Lebanon_Medieval’s ancestry when ancient East Africans are used in the admixture model (Table S10). The final genetic change observed in Lebanon occurred after the Crusaders’ period but, as we showed previously,4 was not related to the Crusaders themselves. We found4 an increase in ancestry related to populations from the Caucasus and Turks in the modern Lebanese population after the medieval period (Figure S11C and Table S11). Using admixture-induced linkage disequilibrium (LD) decay,47,48 we show that admixture occurred around 1640–1740 CE when Lebanon was under Ottoman rule (Figure S12). The LD-decay test also detects significant admixture that occurred during the Hellenistic period, which is consistent with our more direct inferences from the ancient individuals analyzed here (Figure S12). Finally, we fit all the ancient and modern Lebanese data into an admixture graph model showing their relationship with other ancient populations by using data in set 2. The graph supports the results reported here, showing substantial genetic continuity in Lebanon since the Bronze Age interrupted by three significant admixture events during the Iron Age, Hellenistic period, and Ottoman period, each contributing 3%–11% of non-local ancestry to the admixed population (Figures 3 and S13). Figure 3. An Admixture Graph Model for Ancient Lebanon A graph model that fits our data showing the relationship between the ancient Lebanon populations and the admixture events that contributed to the population until modern times. Worst f4-statistics, Iran_N,Levant_N;EHG,Qed_LRoman; Z score = 3.0. See Figure S13 for alternative graph models. In this study, we present new whole-genome sequence data from ancient individuals who lived in the Near East between the Iron Age and the Roman period, spanning a time marked by major historical events and population movements. Our data capture the genetic outcome of some of these events but also show that the genetic composition of the general population was minimally affected and that great cultural transitions in the Near East were not in these cases matched by comparable genetic transitions. Yet, we show that the small genetic changes we detect when using ancient populations sampled from a time series have the power to provide information about past events with details that complement the available historical records.
|
|