April 8, 2018

Luxmanda: a 3,000 years-old proto-Horner in Tanzania

I knew, more intuitively than rationally, that the Horner (Ethiopian, Somali, Eritrean) type of Afro-Eurasian admixture was very old but no idea it was so much. I knew that West Eurasian Upper Paleolithic had an impact on Africa (LSA) but I did not know it went so deep South nor that it had left such a massive legacy as ancient DNA reveals.

Pontus Skoglund et al. Reconstructing Prehistoric African Population Structure. Cell 2017 (open access). DOI:10.1016/j.cell.2017.08.049

The data analysis speaks volumes by itself:

Figure 1 - Overview of Ancient Genomes and African Population Structure

Figure S2 - Ancient Individuals and African Population Structure

Figure 2 - Ancestral Components in Eastern and Southern Africa

We show bar plots with the proportions inferred for the best model for each target population. We used a model that inferred the ancestry of each target population as 1-source, 2-source, or 3-source mixture of a set of potential source populations.

So much that I don't really know what else to say. Of course this is just a sample of what there is in the paper, read it. I'm sure there will be plenty of comments even if the study was published months ago.

Regarding haploid DNA, I don't see anything outstanding but, as I know there is generally quite a bit of interest, these are screenshots of the ancient lineages found (full data in the supp. materials of the study):

Ancient Y-DNA (screenshot)
Ancient mtDNA (screenshot)

Related: No Iberian in Iberomaurusian.

Correction: I first titled this "a 30,000 years-old...". That was a major error on my part and I apologize for any confusion it may have caused. Thanks to Capra Internetensis for spotting it.


  1. major typo in your title there - 3 000 years not 30 000. sure would be nice to have 30 000 year old aDNA from East Africa though!

    1. WTF. You're absolutely right. I knew it, I read 3000 the first time I visited the paper but this morning I somehow read 31,000 what made me blink but still not blink enough.

      I'll correct immediately and thank you very much for spotting that MAJOR error (much bigger than a "typo").

  2. Kaixo Maju
    The luxmanda results very interesting.
    As you know in east africa there is a sandawe people which are a khoisan pocket
    Also at 10.000 BC around the nile delta has oldest massacre grave and victims probable from khoisan race.
    The iberomaurasian article mentioned that ancient Rift people mixied with a race which look Like khoisan.
    I wonder before early holocone -what was also Sahara regrassing- all africa except Western and Atlantc coastal rainforest zone were peopling by khoisan people.
    Do you know something or what do you think
    Gau ona.

    1. The Sandawe are click-language speakers but not Khoisan in any meaningful way: not linguisticallt, not genetically. They are distantly related but only very very distantly so. They rather represent one of various East African ancient populations.

    2. "Also at 10.000 BC around the nile delta has oldest massacre grave and victims probable from khoisan race."

      Can you provide a source? I would expect by that time that ALL North Africa was already more or less "asianized", because the arrival of Upper Paleolithic was much older (50-40 Ka BP, much like in Europe, Central Asia, etc.)

  3. https://www.google.com.tr/amp/s/www.independent.co.uk/news/science/archaeology/saharan-remains-may-be-evidence-of-first-race-war-13000-years-ago-9603632.html%3famp

    "Cranial analysis of the Jebel Sahaba fossils found that they shared osteological affinities with a hominid series from Wadi Halfa in Sudan.[6] Additionally, comparison of the limb proportions of the Jebel Sahaba skeletal remains with those of various ancient and recent series indicated that they were most similar in body shape to the examined modern populations from Sub-Saharan Africa (viz. 19th century fossils belonging to the San population, 19th century West Africa fossils, 19th and 20th century Pygmy fossils, and mid-20th century fossils culled from Kenya and Uganda in East Africa). However, the Jebel Sahaba specimens were post-cranially distinct from the Iberomaurusians and other coeval series from North Africa, and they were also morphologically remote from later Nubia skeletal series and from fossils belonging to the Mesolithic Natufian culture of the Levant.[7]" from Wikipedia

    1. The article says some very different things from the ones you stated earlier: (1) it's not at the Nile Delta but in Northern Sudan, i.e. south of the Second Cataract; (2) it says they were "broadly Subsaharan", hence not clearly Khoisanid in any way, just not "Caucasoid". Also calling it a "race war" is absolutely speculative because we have no idea who the attackers were.

  4. Agur Maju
    If you are following, you know that ancient DNA's related to Iran and Central Asia are being phenomena in eurogenes recently.

    This is a article about the Ganj Dareh early neolithic archeology zone in the central Zagros.


    This paper was interest me due to it seem Zagros neolithic was very close to CHG.
    For me that is a a evidence of İranian origin of Yamnaya culture,as you know half of the ancestry of yamnaya was coming from an iranian source which was look Like CHG,even this both source-iran and CHG-resemble each other.
    Also genetic workings showed that yamnaya people had Olive skin,Brown eyes and dark hairs very similar to modern iranian people:pashtuns,farsi,kurds etc.
    I think ultimate origin of indo european coming from that Zagros neolithic because if they came from caucasus they Must be liked white skinned caucasian people not iranian and oldest steppe culture in europe,khalivinsk emerged in the Volga valley near the Kazakistan border not in the kalmykia,the caucasian border.
    That is my opinion,what is yours.

    1. I don't recall having read/discussed that one but around the same time Lazaridis et al. also published on various genomes of Neolithic West Asia with a similar result for what they call "Iran Neolithic" (Zagros): https://forwhattheywereweare.blogspot.com/2016/06/ancient-genomes-from-neolithic-west-asia.html

  5. https://docs.google.com/spreadsheets/d/1Rja6ZyjQrz3UK7_HTagkzczoOFjwcXOGdNOm9FC3Rik/edit?usp=sharing

    Good afternoon Maju
    I saw this document in eurogenes,it has all steppe related Y-DNA with all SNP's
    You should look this.

  6. http://www.pnas.org/content/111/25/9211

    Agur Maju

    I do not know if you have seen this before but this work was both enlight me and confused.
    Enlighted,because this study is explaining how oldest european neolithic settlements appeared in south greece and albania not in thrace
    Confused,because basque is very outlier from other european people's genetic structure,I was think that basques came from iberian neolithic via cardium pottery but basques does not same as sardinian and south italian
    I will be happy if you explain this basque sitation and good readings.

    1. I had not seen it before but I don't see what's exactly the point and how they "demonstrate" it either. By now it seems quite apparent that West Anatolian Neolithic was crucial in any case.

      Iberians come from ancient (Bronze Age) Iberians who were Basque-like (and Scandinavian-Neolithic-like) and who in turn replaced even more ancient Iberians who were Sardinian-like. Modern Iberians seem to be a mix of those Basque-like Bronze Age Iberians and Indoeuropeans arriving via France (Celts) and Italy (Romans). See this: https://forwhattheywereweare.blogspot.com/2018/03/quickie-pre-indoeuropean-evolution-if.html

      That study is not considering the effect of the Kurgan or Indoeuropean expansion in the modern make-up of Europeans. Basques and Sardinians, lacking that element, are naturally distinct. Sardinians being insular even more clearly so.

      In supp. fig. 9 for example you see Sardinians very isolated (as they are the only "living fossil" of Early Neolithic) but with thin lines linking them to Italy and SE Europe. Basques are also isolated but less so and link somewhat strongly to France and Ireland (no Iberian samples) and weakly to Italy and Denmark. That makes the Basque-Sardinian relation almost compulsorily via Italy, what makes all sense if we understand it is a very old and from the Neolithic (some 7000 years old), however possible Basque-Sardinian links via Iberia are unexplored in the paper you linked to.

  7. https://www.google.com.tr/url?sa=t&source=web&rct=j&url=http://rspb.royalsocietypublishing.org/content/284/1867/20171540.full.pdf&ved=2ahUKEwjx8JnzmszaAhXP0aYKHUwyALs4FBAWMAV6BAgGEAE&usg=AOvVaw2SpSiDTFKtemjoZ2ysSm0g

    Gau ona Maju
    This is a fresh(some six-seven months old)work of globular amphora culture genetic structure and affinity to PIE yamnaya
    As you know some steppist assert that globular amphora was an early wave of yamnaya and responsible for creating of corded ware
    But this work break their dreams

    1. Gabon ba, Mem.

      "Gabon" (just like the African country) is how it is actually said, it is a contraction of "gau on", the nominative final -a is not used (just as you don't say "the good night" in English). Similarly it's said "egun on" (good day = good morning) and "arratsalde on" (good afternoon/evening-before-sunset).

      Anyway, I get an "error 404 - not found" from Google for this link: "http://rspb.royalsocietypublishing.org/content/284/1867/20171540.full.pdf". However if I paste this link, I do get to it.

      Conclusion, do not use google-mediated links but direct links.

      Let's see what is all about. Remember that a recent study on Balcanic Neolithic produced a Globular Amphorae sample that was quite clearly pre-Indoeuropean: https://forwhattheywereweare.blogspot.com/2018/03/ancient-dna-from-balkans.html

    2. Well, that: Globular Amphorae is definitely pre-Indoeuropean and (some) archaeologists did not interpret correctly the tumular elements in the precursor Baalberge culture (which is within Funnelbeaker greater cultural area, probably derivde from Denmark).

      It's clear now that the Indoeuropean expansion into Central and Northern Europe was very sudden and corresponds mostly with Corded Ware, which is almost like the Vedic Era in India: looking all very rural-herder and not at all "civilized".

      Prior to that the inhabitants were roughly Basque-like, although I'd imagine they rather carried R1b-U106, at least in some cases.

      The most curious thing to me is the green component in K=4, because it seems to establish a pattern of continuity between some Vasconic Neolithic peoples and not others: Els Trocs (Catalonia EN) > La Mina (Soria MN) > Iberia Chalcolithic & Globular Amphorae, but not Sweden MN in spite of the similitude between levels of WHG-ish genetics at K=3.

      There's some hidden complexity that we are only beginning to unravel in Vasconic Neolithic. It will take time though to clarify it well enough.

      Pity there is no Y-DNA, because it could shed light on some issues.

    3. It's an interesting read. Not sure if I'd be able to discuss it in any focused manner though: lots of analyses but nothing too specific in the results.


Please, be reasonably respectful when making comments. I do not tolerate in particular sexism, racism nor homophobia. Personal attacks, manipulation and trolling are also very much unwelcome here.The author reserves the right to delete any abusive comment.

Preliminary comment moderation is... ON (it may take some time before your comment is published, thank that to Trumptards).