Origin of the AASI lineage and its specific regional substructure

Jacob Harringer
11 min readFeb 21, 2024

--

The proposed AASI lineage, which is hypothesized to represent the ancestry of the very first hunter-gatherers and peoples of the Indian subcontinent, formed around ~40,000 years BCe after having diverged from a wider Ancient East Eurasian meta-population during the Initial Upper Paleolithic (IUP) period (starting at >45kya), which also gave rise to ‘Australasians’ and ‘East/Southeast Asian people’.

Shinde et al. 2019 noted that both Andamanese Onge or East Siberian groups can be used as proxy for the non-West Eurasian-related component in the “qpAdm” admixture-modelling of AASI, because both populations “have the same phylogenetic relationship to the non-West Eurasian-related component, likely due to shared ancestry deeply in time”. According to Yang (2022), the distinct South Asian ancestry, denoted as the Ancient Ancestral South Indian (AASI) lineage, was only found in ancient and present-day South Asians. Present-day Onge from the Andamanese Islands are the best reference population to date, but Narasimhan et al. used qpGraph to show that the divergence between the AASI lineage and the ancestry found in present-day Onge was very deep. Yelmen et al. 2019 noted that the AASI diverged from the ancestor of both East Asians and Andamanese, and proposed the Southern Indian tribal groups, such as Paniya and Irula as better proxies for indigenous South Asian (AASI) ancestry.

Based on genetic distance and divergence pattern, the AASI are defined as own diverged branch of the wider Ancient East Eurasian meta-population:

The term Ancient East Eurasian, alternatively also known as East Eurasian or Eastern Eurasian, is used in population genomics to describe the genetic ancestry and phylogenetic relationship of diverse populations primarily living in the Asia-Pacific region, belonging to the “Eastern Eurasian clade” of human genetic diversity, and which can be associated with the Initial Upper Paleolithic (IUP) wave, following the Out of Africa migration. Deep East Eurasian lineages, such as Ust’Ishim, Bacho Kiro, and Oase, diverged from the rest of East Eurasians quite early, using different routes, but subsequently becoming extinct or absorbed by West Eurasians.

A single major migration of modern humans into the continents of Asia and Sahul was strongly supported by earlier studies using mitochondrial DNA, the non-recombining portion of Y chromosomes, and autosomal SNP data [42–45]. Ancestral Ancient South Indians with no West Eurasian relatedness, East Asians, Onge (Andamanese hunter–gatherers) and Papuans all derive in a short evolutionary time from the eastward dispersal of an out-of-Africa population [46,47]. The HUGO (Human Genome Organization) Pan-Asian SNP consortium [44] investigated haplotype diversity within present-day Asian populations and found a strong correlation with latitude, with diversity decreasing from south to north. The correlation continues to hold when only mainland Southeast Asian and East Asian populations are considered, and is perhaps attributable to a serial founder effect [50]. These observations are consistent with the view that soon after the single eastward migration of modern humans, East Asians diverged in southern East Asia and dispersed northward across the continent.

Major East Eurasian ancestry lineages which contributed to modern human populations include the following:

  • Australasian lineage — refers to an ancestral population that primarily contributed to human populations in a region consisting of Australia, Papua, New Zealand, neighboring islands in the South Pacific Ocean and parts of the Philippines. Represented by present-day Australasians, e.g. Papuans and Aboriginal Australians, as well as the Philippine Negritos.
  • Ancient Ancestral South Indian lineage — refers to an ancestral population that primarily contributed to Indigenous South Asians. Partially represented by 5,000–1,500 year old Indus Periphery individuals as well as modern South Asians. Highest presence among tribal groups of southern India like the Paniya and Irula. While the lineage is occasionally represented by the distantly related Andamanese peoples, serving as an imperfect proxy, the Andamanese groups are genetically closer to the ‘Basal East Asian’ Tianyuan man.
  • East and Southeast Asian lineage — refers to an ancestral population that primarily contributed to humans living in East and Southeast Asia, much of Remote Oceania, as well as Siberia and the Americas. Represented by ancient Tianyuan and Hoabinhian specimens and present-day East and Southeast Asians.

The Australasian, Ancient Ancestral South Indian, and East and Southeast Asian lineages display a closer genetic relationship to each other than to any non-Asian lineages, and together represent the main branches of “Asian-related ancestry”, which diverged from each other >40kya.

As such, we now know that the indigenous South Asians (AASI) formed their own branch of the wider East Eurasian lineage, and represent the South Asian-specific ancestry component.

Now we will take a closer look at the AASI lineage and its internal diversity and contact events:

The AASI lineage forms a significant ancestral source for modern South Asian populations in terms of full genome and autosomal DNA makeup.

In terms of uniparental haplogroups, the AASI contribution to modern South Asians is less clear, but we have some hints:

Y-chromosome haplogroups which may be associated with AASI-like ancestry are the clades:

  • H1 and H3
  • K (K1 and K2)
  • and C1

The split of IJK* between IJ and K may be linked to the divergence of early West and East Eurasians, shortly before their dispersal waves from the Middle East.

As K clades are nearly exclusively found among East Eurasians, deeply nested withing the AASI, Australasian, and ESEA groups, as well as among IUP remains, but absent from UP West Eurasians, it is plausible to see the local South Asian K subclades as possible AASI-affilated haplogroups, together with local C and D1 clades which were part of the ENA haplotype diversity (eg. C, D1, K branching off with Ancient East Eurasians).

Another possible candidate is haplogroup H, specifically the subclades of H1 and H3. H itself may have originated before the split of East and West Eurasians, as H2 is found mostly, althought at low frequency, among West Asians/Levantines and Europeans. Similarly, C1 is found in both West and East Eurasians, while C2 is solely found among ESEA-derived groups. This suggests that H and C were shared by the last common ancestor of West and East Eurasians some 50–55kya, together with IJK* and D1.

A possible correlation between haplogroups and autosomal DNA can be observed by looking at AASI-rich tribal groups. Here we find C1, K (both K1 and other K clades), and H1/H3 to be significantly higher than among caste and high caste populations. This may support either an AASI-affilation or later bottleneck events among them. In this regard, we will have to wait for actual AASI remains to be discovered and analysed, to know which haplogroups they carried. IMO a plausible candidate would be K and C1 clades and maybe H1/H3(xH2).

Note: It should be clear that haplogroups are not a good indicator of genetic divergence and overall ancestry. They make up a very small part of our total genome and can underwent strong drift and bottleneck events among different populations or groups. As such, carrying a specific haplogroup does not indicate a genetic link to a population with a high frequency of that haplogroup.

Mt-DNA haplogroups associated with the AASI are mainly subclades of the Asian macro-haplogroup M. There are multiple subgroups. Like in the case of Y-DNA haplogroups, the Mt-DNA haplogroup does not give much information on the overall ancestry.

IUP expansion associated with the dispersal of Ancient East Eurasians, and UP expansion, associated with the dispersal of Ancient West Eurasians.
Possible migration route of Y-chromosome K clade and subclades; South Asian tribal groups display an increased amout of K1* and K2* clades.

Now let us come to their internal diversity and contact with surrounding lineages:

We are quite sure now that the AASI displayed a substructured genetic makeup, and that they broadly could be differentiated into AASI_NW (Northwest) and AASI_SE (Southeast).

The Northern or Northwestern AASI are argued to be the main source of AASI ancestry for most modern South Asians viz a viz the Indus Valley civilization (IVC):

The IVC formed by the merger of Eastern Iranian hunter-gatherers, distantly related to the Neolithic Iranian/Zagrosian farmers from Western Iran, with local Northwestern AASI (AASI_NW):

Later on, IVC ancestry expanded throughout South Asia, merging with local tribes to give rise to the ASI cline, while remaining IVC ancestry in the North absorbed the arriving Steppe pastoralists (Indo-Aryans) to give rise to the ANI cline, together forming the ANI-ASI or Indian cline known from modern South Asians.

The difference between AASI_North and AASI_South is not only evident in this demographic event, but also by comparing AASI_North ancestry modeled out from Northern Indian’s non-Western ancestry, and AASI_South ancestry modeled out from Southern Indian tribal’s non-Western ancestry. We get quite interesting results:

While AASI_North seems to represent a purely Proto-AASI-derived ancestry (with no Western inputs), the AASI_South display evidence of geneflow from Southeast Asian hunter-gatherers (Basal East Asians aka Hoabinhians); The same holds true when replacing Hoabinhians with the Tianyuan lineage:

Based on that, it is very likely that there was some kind of Hoabinhian/Tianyuan-like geneflow into South Asia, at least southern and eastern India. This is evident by the relative higher affinity for Hoabinhian/Tianyuan for southern and eastern Indian tribals groups, when compared to Northern groups regardless of their AASI ancestry.

Some people already noted that AASI_S as well as the Paniya display an odd Hoabinhian/Tianyuan affinity absent from more northern groups, the likely reason for this is indeed geneflow from Southeastern Asia before any Western ancestries arrived to South Asia, or at least Southern India.

Sometime before the arrival of Belt Cave/Hotu people, there was an east-to-west migration of LAO_Hoabinhian or at least a very similar population. As the same suggests, these people lived in Laos during the early neolithic time period, around 7888 ± 40 BP. Genetically, they are very similar to the Andamanese and the Paleolithic Tianyuan man from Northern China, and are considered to be “basal east asian”. LAO_Hoabinhian migrated to South Asia and mixed with the preexisting AASI populations there — but only to some degree.

AASI and Lao_Hoabinhian are frequently confused with one another. The most classic example of this is our perspective on the Paniya and Irula people. Originally these south indian populations were thought to be upwards of 80% “pure” AASI. But in the last few years, it has become abundantly clear that they carry significant ancestry from LAO_Hoabinhian. So in reality, Paniya has over 20% “basal east asian” or Andaman-like ancestry, and less than 60% “pure” AASI ancestry. Looking at east indian Munda-speakers, we also see 10–15% LAO_Hoabinhian in addition to AASI and more recent bronze age Austroasiatic ancestry.

As example, we can look at the genetic makeup of some groups and compare their relative AASI_North and AASI_South ancestries, and there Hoabinhian/Tianyuan component:

Based on the above, a correlation becomes quite clear. Furthermore, the Odisha_Dhurwa tribe in Odisha, Eastern India, has an even higher Tianyuan/Hoabinhian component than the Irula or Paniya (via AASI_S and extra), which suggests that there was also an Eastern variant of the AASI, an AASI_East, which may have been primarily Hoabinhian/Tianyuan-like.

All this points to geneflow from an Hoabinhian/Tianyuan-like source from Southeast Asia into Southern Asia, resulting in the formation of AASI_E and AASI_S, while AASI_N did not receive such input.

This finding is corroborated by some papers which differentiated different types of AASI ancestry, such as in this model:

And studies which noted that there is evidence from population genomics and archaeologic data on sea level rise, for a quite significant migration of Southeast Asian hunter-gatherers into Southern Asia:

IMO, this explains the Hoabinhian/Tianyuan affinity for southern and eastern Indian tribal groups and the substructure of the AASI into AASI_N, AASI_S, and AASI_E.

This affinity is also evident in their genetic distance:

Overall, the AASI form a primary branch of Ancient East Eurasians, next to Australasians and East/Southeast Asians. The AASI_S is significantly closer to the ESEA branch, specifically via Hoabinhian-like geneflow from Southeast Asia, while AASI_N does not seem to have any “non-AASI” component and is derived entirely from the local East Eurasian variant.

At a speculative level, the phenotyps associated with the proper AASI may be somewhat like this one:

For an interesting comparison, we can look at the reconstruction of Oase 2, a Basal East Eurasian IUP sample (c. 42kya), which is forming a sister lineage to the ancestor of AASI, Australasians, and Eastern Asians:

Oase 2, IUP specimen sharing affinity for Ust’Ishim and Tianyuan remains and modern East Eurasians; his specific lineage however went extinct.

AASI-like ancestry among Iranian Mesolithic and Neolithic remains:

There is some evidence that beyond the primarily Basal Eurasian and UP European/West Eurasian components among Iran_N and Iran_Meso, there was also a small AASI-like component (or a deeper ENA component). This may be related to K1 (LT), or just later mutual geneflow from South Asia (eg. Eastern Iranian HGs spreaded into South Asia, and South Asian HGs westwards — or via a cultural-economic networks and contacts).

So far, some models show ENA components of around 10% for Iran_N:

ENA represented by Onge, additional Basal Eurasian is simulated by Mbuti
qpGraph on Ganj_Dareh_N (hypothetical)

For an alternative model, see Allentoft et al. 2024:

Here, ENA components are indirectly derived via the ANE; 52% ANE = 18,2% Tianyuan-like.

Eg. Iran_N and Iran_Meso may have an additional, albeit small, AASI-like component, but it may also just be indirect ENA via the ANE, which had around 35% Tianyuan-like ancestry, with the rest being Basal Eurasian and UP European-like.

Note: Basal Eurasian is a proposed lineage of anatomically modern humans with reduced, or zero, archaic hominin (Neanderthal) admixture compared to other ancient non-Africans. Basal Eurasians represent a sister lineage to other Eurasians and may have originated from the Southern Middle East, specifically the Arabian peninsula, or North Africa, and are said to have contributed ancestry to various West Eurasian, South Asian, and Central Asian as well as African groups. Ferreira et al. in 2021 argued for a point of origin for Basal Eurasians into the Middle East, specifically in the Persian Gulf region on the Arab peninsula. Among modern populations, Basal Eurasian ancestry peaks among Arabs (such as Qataris) at c. 45%, and among Iranian populations at c. 35%, and is also found in significant amounts among modern Northern Africans

Conclusion

The AASI lineage is one primary branch of Ancient East Eurasians, next to Australasians and East/Southeast Asians (ESEA). The AASI_South (and East) received geneflow from Basal East Asian (Hoabinhian-like) hunter-gatherers from Southeast Asia, while the AASI_North do not seem to have any non-AASI input.

The AASI ancestry is significant for all modern South Asian derived groups and is their specific genetic variant out of the wider human genetic diversity.

I hope this post was interesting and clarified some points of the AASI, their origin, and their affinities.

Thank you! Jacob

--

--

Jacob Harringer
Jacob Harringer

Written by Jacob Harringer

Hello, I am Jacob Harringer. I am very interested in human history and population genomics. I love to read academic papers and to travel.

Responses (1)