A more recent, revised and updated version of this paper has been published (2019)

Admixture analysis

From Indo-European.info

Revision as of 18:22, 30 May 2017 by Admin (talk | contribs) (Created page with "Genetic admixture refers to the analysis of the gene flow between populations that had previously been relatively isolated from one another. Since isolated populations develop...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

Genetic admixture refers to the analysis of the gene flow between populations that had previously been relatively isolated from one another. Since isolated populations develop linguistic differences relatively quickly, linguistic changes might be expected in a newly hybridised population (Jobling et al. 2014). However, pidgin languages are quite rare, and often one language – usually that of the successful migrants – becomes the superstrate, and another the substrate.

On the other hand, language and culture are unlike a genome in several different ways. While it is possible to obtain admixture percentage of any ancestral population, ancestral language reconstruction and its identification with cultures needs the intervention of careful anthropological investigation. For admixture results to be meaningful, studied loci have to be correctly averaged (and samples should be as complete as possible); genetic drift and selection since admixture have to be taken into account (e.g. distant populations might show a higher differentiation from the original territory); ancestral populations have to be correctly identified, including their number and precise alleles (Jobling et al. 2014). Ancient DNA is best collected with the goal of testing specific hypotheses.

Some linguists have used the biological foundations of phylogenetics to extrapolate questionable methods to linguistics, and have thus obtained questionable results (Gray and Atkinson 2003). Similarly, scientists are using the available statistical means to study genetic admixture in modern human populations, extrapolating admixture mapping methods to scarce ancient human samples, and deriving simplistic, far-fetched conclusions. This paper demonstrates the need to include wide anthropological investigation of the historical context of the samples studied, including linguistics, archaeology, and cultural anthropology, as well as careful investigation of haplogroups, to obtain plausible explanations for the complex data obtained in human biology.

It has been proposed that migrating Yamna pastoralists into already expanding Corded Ware groups (Wencel 2015) might have created the necessary environment for the spread of Indo-European languages. Previous mainstream models for Indo-European expansion, based on the “kurgan hypothesis” (Gimbutas 1977) associated the spread of Pre-Germanic (adopted on the Dniester) and Pre-Balto-Slavic (adopted on the middle Dnieper) to the expansion of Corded Ware cultures (Anthony 2007). Given the lack of direct cultural connections between Yamna and the Corded Ware culture, this spread was explained in terms of either an incorporation of languages through centuries of interaction into Funnel Beaker cultures, or through the emulation of the language of Indo-European chiefs by Corded Ware cultures (beginning ca. 2700-2600 BC) for politico-religious reasons (Anthony 2007).

The admixture of Yamna aDNA samples found elevated (up to 76%) in Corded Ware samples has been said to support the migration of Yamna populations into Corded Ware groups, while the lower percentage found in Bell Beaker and Únětice groups (50-70%) has been explained as a subsequent, less profound displacement process triggered by western and central European groups (Haak et al. 2015; Allentoft et al. 2015; Mathieson et al. 2015). It has also been found that samples from Globular Amphorae culture do not show evidence of steppe ancestry (Mathieson et al. 2017).

These limited results, apparently challenging archaeological interpretations previously considered established, are propagating quickly within the field of Indo-European studies. David W. Anthony has recently supported the appearance of the Corded Ware culture through the contacts of Yamna immigrants with indigenous people of the Globular Amphorae culture in southern Poland (Anthony and Brown 2017), based on their previously known contacts and early dating. Similarly, Kristian Kristiansen has supported the dominance of Corded Ware in central Europe south and north of the Carpathians, asserting that their pottery was apparently shared later by the Bell Beaker culture (Kristiansen et al. 2017).

Many concerns have been raised about obtaining simplistic conclusions based on genetic results (Heyd 2017):

  • Samples from the Pontic-Caspian steppe – from which ‘steppe ancestry’ is defined as a precise combination of West hunger-gatherer and Caucasian hunter gatherer ancestry – are scarce, and recent ones only from one eastern region (Kalmykia).
  • Steppe ancestry has been found in Corded Ware, Afanasevo, Andronovo, and Srubna cultures, all of which – and even a late individual of Bronze Age Bulgaria from Merichleri , ca. 1690 BC – show higher steppe ancestry than samples clearly identified as from Yamna migrants in the Balkans. Samples from central Balkans show in fact a relative increase in steppe ancestry later, during the Bronze Age (unlike west Europe and the southern Balkans).
  • The north-west Pontic area – from where many Yamna migrants seem to have expanded west along the Danube – had been a zone of interaction with peoples from the Danube and the Eastern Baltic for millennia – and could thus cluster closer genetically to peoples from Carpathian cultures than the eastern Pontic-Caspian steppe. Peoples from this eastern zone, whose samples are used to define steppe ancestry, had migrated to the north-west Pontic area at least twice, first in the formation process of the early Khvalynsk-Sredni Stog cultures, and later in the formation of the Yamna culture (see above), probably creating a more mixed western genetic pool, more similar to the one found in western Yamna migrants.
  • Scattered samples from different periods (by millennia) from the Forest Zone and steppe already showed certain common clusters before the Neolithic and Chalcolithic expansions in global ancestry profiles, in the first articles published . More recently, Estonian samples have shown a genetic component associated with Caucasus hunter-gatherers coinciding with the spread of R1a-Z645, which rules out Corded Ware and Yamna as the only origin of this component (Saag et al. 2017). Supporting these conclusions, a sample of a female from Zvejnieki, dated ca. 2885 BC and classified as from Latvian Neolithic cultures (Jones et al. 2017), has been found to cluster closely with Yamna samples (Mathieson et al. 2017), closer than samples from Corded Ware cultures, and before any described migration from Yamna could have happened to this region. These recent samples question the validity of assuming a direct gene flow from Yamna migrants when explaining the so-called steppe ancestry found in samples from Corded Ware cultures.
  • Two female samples from Bohemia were misidentified as Bell Beaker (Allentoft et al. 2015), when they were in fact three millennia younger, from Czech Slavs (Mathieson et al. 2017). PCA did not (and cannot) show differences with Bell Beaker or Balkan samples, since parental populations need to be available, or else archaeological context is needed to define demographic models and potential ancestral populations, to ascertain their actual link to the so-called steppe ancestry.
  • Corded Ware samples are late, almost coinciding with the Bell Beaker expansion. No samples have been published from potentially controversial areas – like the Contact Zone, eastern Baltic and western Yamna – during the most relevant periods. Old samples (closer to admixture events) tend to show a higher range of variation, and could inform better of the real impact of migrations, while younger ones – depending on non-random mating processes, influenced by geographic structure or socioeconomic factors – may falsely show a relatively homogenous high or low ancestral contribution (Jobling et al. 2014).
  • The migration of Pontic-Caspian steppe into Neolithic/Bronze Age Central Europeans has been argued to be strongly male-biased (Poznik et al. 2016), with a study suggesting up to 14 migrating males for every migrating female (Goldberg, Günther, et al. 2017) , but different in the rates regarding Corded Ware, Bell Beaker, and Únětice. The results of the latter study have been disputed (Lazaridis and Reich 2017), and this in turn contested by the original authors based on the impact of small, low-coverage ancient samples in admixture analyses (Goldberg, Gunther, et al. 2017).
  • Ascertaining differences in demographic changes is especially important in light of an apparently mostly peaceful Yamna migration along the Danube (Heyd 2012), contrasting with the potentially violent and strong patrilocality shown by peoples of the Corded Ware cultures (Kristiansen et al. 2017). Also relevant is then the actual increase in population due to such expansions – greater in south-east Europe (Müller 2013) –, which further influences the genetic drift observed.
  • In the academic community prestige, access to grants, and even jobs depend on getting articles published in journals of high impact factor. These journals prefer short articles, mainly based on mathematical methods (preferably with reference to improvements in such methods), groundbreaking conclusions, and self-important titles, with a tendency to “culture-historicism”. This trend is very well represented and remembered in anthropological disciplines by Nature’s paper on the Anatolian origin of Proto-Indo-European, based on glottochronology, by Gray and Atkinson (2013).
  • SNP investigation offers a simple view of one’s own paternal line, that a thousand years (or ca. 30 generations) ago would represent a 1,000,000,000th of one’s own genealogical tree; four or five thousand years ago, its contribution to a personal ethnic definition is almost non-existent. This, together with the perceived complexity (and lack of familiarity with) intricately linked anthropological disciplines, has made admixture analysis quite popular among amateur geneticists, who can easily play with published open source software programs, due to their accessibility. However, the correct use of these programs needs much more than just knowing how to apply certain commands to some data. The quest for one’s own personal and national “ethnic proportion”, often as part of pre-existing simplistic ethnic beliefs and socio-political agendas, is also being promoted by commercial genetic testing companies to sell their products, in what would certainly be a reason for Kosinna’s smile today.