More Neanderthal admixture details

16 Feb
The main conclusion of this new study seems to be that East Asians have normally more Neanderthal admixture than South or West Eurasians. Also the Maasai from East Africa have been shown to have minor Neanderthal admixture, consistent with the also minor Eurasian genetic component they have.
Jeffrey D. Wall et al., Higher Levels of Neanderthal Ancestry in East Asians Than in Europeans. Genetics 2013. Freely accessible at the time of writing thisLINK [doi: 10.1534/genetics.112.148213]


Neanderthals were a group of archaic hominins that occupied most of Europe and parts of Western Asia from roughly 30-300 thousand years ago (Kya). They coexisted with modern humans during part of this time. Previous genetic analyses that compared a draft sequence of the Neanderthal genome with genomes of several modern humans concluded that Neanderthals made a small (1-4%) contribution to the gene pools of all non-African populations. This observation was consistent with a single episode of admixture from Neanderthals into the ancestors of all non-Africans when the two groups coexisted in the Middle East 50-80 Kya. We examined the relationship between Neanderthals and modern humans in greater detail by applying two complementary methods to the published draft Neanderthal genome and an expanded set of high-coverage modern human genome sequences. We find that, consistent with the recent finding of Meyer et al. (2012), Neanderthals contributed more DNA to modern East Asians than to modern Europeans. Furthermore we find that the Maasai of East Africa have a small but significant fraction of Neanderthal DNA. Because our analysis is of several genomic samples from each modern human population considered, we are able to document the extent of variation in Neanderthal ancestry within and among populations. Our results combined with those previously published show that a more complex model of admixture between Neanderthals and modern humans is necessary to account for the different levels of Neanderthal ancestry among human populations. In particular, at least some Neanderthal-modern human admixture must postdate the separation of the ancestors of modern European and modern East Asian populations.

There is a sentence in page 8 in which the authors claim that:

By using the high coverage Denisova genome, we are able to show that the admixture rate into East Asians is 40% higher than into Europeans.

However I fail to see it in the D-statistic graphs, which show less than 15% more Neanderthal admixture in Eastern Asians than in Europeans (on average and discounting error margins, which strongly overlap). It may be therefore another case of hair-splitting.
Instead the evidence of weak Neanderthal admixture (via minor Eurasian  genetic influence) in the Maasai of East Africa seems more solid: while error margins of Maasai and Luhya do overlap, the main blocs do not.

Figure 3. Summary of significance tests for average values of D. Positive values indicate that the second sequence is more similar to the Neanderthal genome than the first sequence. In all parts, the box plots indicate the range of D values obtained for pairs of individuals from the populations indicated. Parts A and B are box plots of individual D statistics computed for each individual from the specified population compared with each Yoruban. The p values are from the randomization test, Test 1, of significant differences in the average D values for different pairs of populations. Parts C and D show box plots of individual D statistics computed for every pair of individuals in the specified populations. The p values are from the randomization test, Test 2, of significant differences of the average D from 0. See also Table 2.

It is surely convenient to look carefully at all the data, including the supplementary materials, before jumping to any strong conclusions. Among them I found more interesting fig. S1:

Figure S1: Box plot of the D-statistics for Analyses A and B for the set (Afr, X), where X was any of the non-African populations,
CEU or TSI (Europeans, green), CHB or JPT (East Asians, blue), or GIH (South Asian, pink) (PDF, 222 KB).

As we can see here, the error margins continue overlapping very strongly, but, assuming that we can jump to conclusions based on the norm (thick black line), which is assuming some risks indeed but also the only way to (tentatively) agree with the authors’ conclusion of greater Neanderthal admixture among East Asians, then the causes should be:
  1. Founder effect among East Asians, known to have less overall genetic diversity and suspected to have undergone some extra (mild?) bottleneck at their ancient origins.
  2. Low-level African (and surely also early OoA residuals from West Asia) admixture among West Eurasians, notably Tuscans (TSI) in these samples but also to some extent all Europeans (incl. CEU). This has the effect of increasing genetic diversity but also of slightly diluting Neanderthal admixture. The control here are Indians (GIH), which show slightly more Neanderthal admixture than Europeans (but are still closer to these than to East Asians also in this aspect).
Indians (who don’t seem to have any post-OoA African admixture, unlike Europeans and West Asians, who have it at variable low levels) suggest that factor #2 (dilution) weights only somewhat and therefore that factor #1 (East Asian founder effect) must be considered the main one instead. 
But, unless someone can point me where I am wrong, I fail to see neither the alleged 40% excess Neanderthal admixture in Orientals nor why would these results question the single admixture episode (or period) at the origins of migration out of Africa (OoA).

To finish this entry, it must be mentioned that the authors could only detect at most the tiniest fraction of Denisovan admixture among the sampled populations (i.e. nothing relevant and with no regional differences). However they did not research, admittedly, the South China populations suggested to have slight more Denisovan input by  Skoglund and Jakobsson 2011.
See also:


61 responses to “More Neanderthal admixture details

  1. terryt

    February 19, 2013 at 2:53 am

    "No". You obviously need to brush up on your knowledge of population genetics then. Autosomal DNA preserves genetic elements older than either mt-DNA or Y-DNA. I would have thought that was obvious to anyone. "Is original from SE Asia and surely not the technology brought by the Andamanese. It's probably more related to the Nicobarese, who speak an Austroasiatic language". Possibly, but the settling of both island groups could be almost contemporary with each other. After all you know as well as I do that technology can be passed to other groups. The boating technology that allowed SE Asian groups to reach the Nicobars would obviously also have passed along the mainland coast allowing south Burmese populations to reach the Andamans. I'm not claiming that is exactly what happened but there is certainly no reason for it to be impossible. "I would prefer to use genetic inference and suggest a very old divergence at unknown but clearly Paleolithic time". But genetic evidence doesn't 'suggest a very old divergence at unknown but clearly Paleolithic time'. Andaman Y-DNA D is considered closely related to Indian Y-DNA D and Andaman mt-DNA Ms both have close relations in India. M32 is even found in Madagascar which was settled no more than some 2000 years ago. "It seems to me still a bit late" Well it does rely on molecular-clockology so I'm prepared to accept an earlier date for SW Asian N's diversification if you so wish. "It is therefore safe, considering the whole and not just cherry picking the evidence, that the earliest UP dates are not older than those from Kara Bom, what leaves us at c. 43 Ka BP, at most 44.9 Ka BP (but could also be more recent, up to 41.7 Ka BP)". You're still assuming, without any evidence whatsoever, that the first modern humans to enter the region carried a fully-developed Upper Paleolithic technology. On what grounds do you make that assumption? "As for Okladnikov cave, further North, it is interesting not for its UP dates, which do not exist, but for its very late MP ones: 48.8-33.5 BP, what makes it a likely last stand for either Neanderthals or Denisovans after displacement by our kin from their more traditional settlements further south". Another unjustified assumption. We know that Neanderthals replaced modern humans in the Levant as climate cooled around 70,000 years ago yet you insist that Central Asian Neanderthals and Denisovans remained exactly where God had placed them until being replaced by modern humans.

  2. terryt

    February 19, 2013 at 3:11 am

    "You obviously need to brush up on your knowledge of population genetics then". Yet you can say: "In addition A0 and A1a are very rare lineages even where they exist and, on top of that, Y-DNA (with no plausible mtDNA counterpart), which is generally much less strongly related to autosomal (overall) genetics than " So what is your difficulty in seeing that Autosomal DNA is presumably usually the oldest with mt-DNA being often younger and Y-DNA being even younger? Or is it just that you can't see it when it suits your purpose to not be able to see it? "On the other side, as Terry points below, this paper does mention that there seems to be some Mousterian, related to that of Altai, in Mongolia. If so, I stand corrected". And you will have even more to stand corrected on as time goes by. "The key point is that the Indian subcontinent was not demographically a single unit in all the last Ice Age". Yet you insist on considering it one unit when considering basal haplogroup diversity in the continent. For example yopu said: "The basal diversity does suggest an origin in South Asia for mt DNA R" So which region within South Asia exhibits this basal diversity? "It was also mentioned in the first source that the Northern Deccan was also arid in the Middle Pleistocene, what affects directly the early population of the subcontinent by our species, which clearly happened long before the late Pleistocene". The paper does say that the Thar Desert is 'arid' and the North Deccan is 'semi-arid'. Any time that the North Deccan was semi-arid it is likely to be a result of the expansion of the Thar desert, making anywhere north of the Deccan uninhabitable except for the upper reaches of the Ganges. That does actually make sense of the apparent centre of dispersal of mt-DNA M though. Several branches were eventually able to move south into Madhya Pradesh while most journeyed east north of the Ganges into Northeast India and Zomia.

  3. Maju

    February 19, 2013 at 8:02 am

    Vale. I'm not familiar with this A00, which is not even listed at ISOGG as of today. I cannot therefore comment further at this stage.

  4. Maju

    February 19, 2013 at 8:13 am

    Neither M31 nor M32 look particularly derived from M, not more than your famous X is from N for example (in the case of M31, quite less in the case of M32). I conclude that they are therefore very old, maybe of 70 Ka ago for M32 and a little less 60-50 Ka for M31. Certainly not "recent". Y-DNA is less obvious to look at but D is not "recent" in any case. "You're still assuming, without any evidence whatsoever, that the first modern humans to enter the region carried a fully-developed Upper Paleolithic technology. On what grounds do you make that assumption?"What?! Sorry I'm talking of Aurignacian-like industries all them mode 4. "Fully developed" is a category I don't understand well in this context. All I say is that they are related techno-cultures of a kind that can be safely associated with H. sapiens in those regions (documented in Palestine and Altai). "Another unjustified assumption".If you say so…We also know that Homo sapiens replaced Neanderthals in the dates discussed here or a bit later in all West Eurasia, what's your problem then? What part don't you understand?

  5. Maju

    February 19, 2013 at 8:15 am

    Read it again.

  6. terryt

    February 20, 2013 at 2:13 am

    "Neither M31 nor M32 look particularly derived from M" Of course they're derived from M. That's why their name starts with 'M'. If you're meaning to say they are not basal M then I agree. That was my point. They are not particularly early in M's diversification. "I conclude that they are therefore very old, maybe of 70 Ka ago for M32 and a little less 60-50 Ka for M31. Certainly not 'recent'". How on earth do you come to that conclusion? Andaman M31 is M31a1, sister to Orissa M31a2. The combined M31a haplogroup is sister to M31b'c. So by the time we get to M31a1 we are considerably downstream from M31. Andaman M32 is M32a, sister to M32c from Madagascar. The combined M32 haplogroup is siter to Indian M56, so again we are dealing with a downstream haplogroup. "not more than your famous X is from N for example" X branches directly off basal N. What are you talking about? "All I say is that they are related techno-cultures of a kind that can be safely associated with H. sapiens in those regions (documented in Palestine and Altai)". So I assume you're prepared to share with us the characteristics that distinguish the technologies of the first modern humans in the Levant from the later Neanderthals that occupied the region. That will be a very useful piece of information because no-one else has yet been able to do so. I agree that mode 4 is a characteristic of Homo sapiens but its absence is not evidence of Homo sapiens absence. "We also know that Homo sapiens replaced Neanderthals in the dates discussed here or a bit later in all West Eurasia, what's your problem then? What part don't you understand?" You are the one who fails to understand that the first humans to emerge from africa did not have a mode 4 technology. Thereforee you cannot use the expansion of mode 4 technology to chronicle Homo sapiens' spread. We also cannot assume that Neanderthals did not replace early modern humans beyond the Levant as climate cooled. A factor to keep in mind here is that Central asian Neanderthals were shifted phenotypically towards Homo sapians as compared with Neanderthals from further west. This is most parsimonioulsy explained as being the product of hybridization between the two 'species'. In fact many individuals have been difficult to classify as either species, only defined by the mt-DNA extracted. For example note the headline of this post:

  7. terryt

    February 20, 2013 at 3:19 am

    "Read it again". So I did, and what I picked up is something you have always dissagreed with. The Petraglia book specifically states that humans prefered the 'deciduous woodland savanna ecosystems'. Bot extracts emphasise the connection between aridity in the north deccan as being part of more widespread climate changes. "This other book by Michael Petraglia talks instead of arid conditions in the last Ice Age in the Western Deccan, suggesting that Eastern and Central India (not affected) was the preferred habitat for human habitation then" But he also says that hominids seasonally exploited even what he calls the 'isolated areas' during arid times. Talking of inhospitable conditions in India: "the founders of East Asian peoples had on average slightly more Neanderthal input than the founders of South/West Eurasian ones. A mere fluke long ago". Doesn't make sense. A route through Central Asia is the only possible explanation for some observed phenomena. For example the absence of any basal mt-DNA N haplogroups in South Asia as well as the lack of any evidence for any technological connection between South and East Asia until somke time after the Tianyuan individual's presence . Here's a paper on human occupation of Central Asia: author does say: "Given that the new dates from the Nihewan basin in northern China place hominin occupation of this region at roughly 1.6 Ma (Zhu et al. 2004) the relatively younger dates from Central Asia are surprising, as the region represents a possible corridor to the east. The observation that paleoclimatic reconstructions support a general similarity between the savanna conditions of Africa, the Levant and southern Eurasia (i.e., Dmanisi) with those of the foothill regions of Central Asia makes the relative absence of Early Pleistocene sites in the latter area even more unexpected." The author is obviously not convinced by the usual explanations for that absence. "it seems hominin groups lived in Central Asia during the coldest anddriest periods of the Late Pleistocene". "Lithic techno-typology as well as hominin morphology (see discussion below) point to Central Asia as a zone of interaction during the Middle Paleolithic, with contacts to the north, east, and west". Note, even to the north. "Evidence that supports potential isolation/separation of Central Asian hominins from northern and eastern Asian populations is not easily delineated. Geographic barriers like the high mountain zones of the Tien Shan and Pamirs did not seem to impede faunal dispersals across Eurasia (seeKeates 2004). In this regard, the notion that Neandertals moved east from their core area in Europe and stopped in Central Asia because of being geographically hemmed in is unsupported, certainly in light of Middle Paleolithic evidence from the Siberian Altai and other sites in China". "In any event, there is some evidence of the continuous occupation of Central Asia during the Late Pleistocene and perhaps across the Middle to Upper Paleolithic divide. A continuous occupation of the Siberian Altai also has been suggested and seems to be supported at sites like Kara Bom (Derevianko et al. 2000; Derevianko and Markin 1992)".

  8. Maju

    February 20, 2013 at 5:07 pm

    Oh great, another post to the trash bin! Consider yourself replied because I did in length and it did not go through.

  9. Maju

    February 20, 2013 at 5:08 pm

    Anyhow you must know well the answers I was going to give to your many polemicist questions. We have been discussing for way too long.

  10. Maju

    February 20, 2013 at 5:11 pm

    I just suggested that the dry Deccan acted as buffer, not barrier, to generate two somewhat different populations in India. They are not that different according to Fst distance values (read previous relevant comment), so maybe it's not that relevant anyhow.

  11. terryt

    February 21, 2013 at 7:19 am

    "Anyhow you must know well the answers I was going to give to your many polemicist questions. We have been discussing for way too long". I agree we 'have been discussing for way too long' but you have consistently ignored the many problems I point out in your belief and concentrate on irrelevant trivial nin-picking.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: