Substack cometh, and lo it is good. (Pricing)

White modern Northern Europeans are genetically more like brown South Asians than brown(ish) ancient Northern Europeans were

The Guardian has a piece by Arathi Prasad, Thanks to Cheddar Man, I feel more comfortable as a brown Briton. Dr. Prasad is a geneticist, so the science is pretty decent (she’s probably seen the documentary ahead of time too).

But there is a curious quirk here and it reveals something about human psychology: modern Britons are genetically much closer to South Asians, like Arathi Prasad, than these ancient darker-skinned Britons. The plot to the left illustrates this (it’s using the Dystruct package). The far right of the top panels represent South Asians. You can see Europeans pretty clearly. Let’s note two things:

1) Modern Europeans (except for Sardinians) share an orange “steppe” component with most South Asians (these are no doubt Indo-European migrations of the Bronze Age)

2) The brown element represents European hunter-gatherers. This element is found at varying quantities across Europe, with the lowest fractions in Sardinians. Though present in South Asians (this may or may not be an artifact to be honest), it’s not present at very high frequencies.

One always has to be careful about taking these proportions as literal representations of ancestral populations. They are not. But what they show is that modern Northern Europeans and South Asians have been touched by the same population movements over the past 5,000 years, and so are genetically much closer than the people who lived in Northern Europe and South Asia 5,000 years ago.

Humans are a visual species. In a pre-modern environment, physical cues were important for group identity, though I suspect just as much due to scarification and tattooing as phenotypic differences due to biology. The fact that Cheddar Man, and Paleolithic hunter-gatherers in Western Europe more generally, probably resembled modern South Asians more than they do modern Northern Europeans (I think they were more likely to be olive-brown than dark-brown, but I’m not confident), is more salient to human folk biology than the fact that modern Northern Europeans are much closer genetically to South Asians than the more “brown” ancient Northern Europeans.

Stuff like this always reminds me of the deep wisdom in Artur C. Clarke’s Childhood’s End. The ultimately benevolent alien species which mentored humanity shielded us from their physical appearance because the knew we’d find it horrifying. The substance of what they did for us, who they were, was going to be less important to immature humans than the fact of what they looked like.

Note: Fst between Sindhi from Pakistan and WHG (Cheddar Man was one) is 0.087. Sindhi from Pakistan and English is 0.023. English to WHG is 0.058 (source). Fst can not be naively interpreted as “genetic distance.” But, this gets at the fact that Mesolithic European hunter-gatherers were very distant from modern South Asians. And widespread gene flow and admixture over the past 5,000 has compressed a lot of genetic differences which were starker across geography in the past.

8 thoughts on “White modern Northern Europeans are genetically more like brown South Asians than brown(ish) ancient Northern Europeans were

  1. OT: I’m getting the following error when trying to view the site occasionally.

    Internal Server Error

    The server encountered an internal error or misconfiguration and was unable to complete your request.

    Please contact the server administrator, [email protected] and inform them of the time the error occurred, and anything you might have done that may have caused the error.

    More information about this error may be available in the server error log.

    Additionally, a 500 Internal Server Error error was encountered while trying to use an ErrorDocument to handle the request.

  2. We have only the Aryan migrations to thank(?) for this proximity.

    If anything, looking at PCAs and ADMIXTURE graphs, Europeans are slowly selecting WHG/SHG/EHG traits again, shifting slowly towards them compared to past populations. CHG ancestry seems to be the most “undesirable” one and is suffering at that selection, slowly decreasing.
    Anatolian Farmer admixture is also being selected against, but recent migratory waves keep pushing it towards their direction.

    Also, Dystruct is crap – modeling Yamnaya like that is idiotic.
    (I posted in the wrong thread before)

  3. Twinkie, i have some ideas why this is happening. will devote much of sunday to this.

    “Guest”, your comment is hard to parse. if you post something that unintelligible in the future i’ll just tag it as spam.

  4. According to Dystruct:

    The Steppe is now descended from a Kostenki like population, it is highest in the Caucasus, expected levels in Europe and South Asia, but now the Middle East has similar levels if not more than Europe and South Asia.

    The dominant component in South Asian populations is now related to Ust’Ishim and Ma’lta, absent in Europe but somewhat higher in the Middle East.

    I’m having a hard time believing this, is there any chance it could be wrong ?

  5. Hard to compare above result with any current analysis of ancient European population genetics, because they are using a 2015 dataset – pre-CHG, and pre-Levant Neolithic and pre-Iran Neolithic. There’s a little more world data today.

    ADMIXTURE represents early steppe as about 50:50 of a European Hunter Gatherer (EHG+WHG) component and a Caucasus/India related component. In rough terms, we know this to telling us something about population history (Yamnaya roughly EHG:CHG in 50:50 terms, India shares a large amount of ancestry with CHG probably via the related Iranian Neolithic).
    Dystruct doesn’t pick this up, but places Yamnaya as constituted of a separate component (which, yes, is the same as K-14’s component). This is most clearly representing a phenomenon in Europe, but less clear in the Caucasus. Of course, this maybe could be “fixed” by running a recent dataset with more references for the Near East, but does raise some questions of how useful the approach is, if adna is somewhat sparse or clustered in a particular region.

    Looking at proportions, Dystruct’s Yamnaya component seems to be correlated with the CHG component in ADMIXTURE; both the CHG component in the ADMIXTURE run and the Yamnaya component in the Dystruct run seem like they would predict the same amount of steppe ancestry in North Europe (assuming all CHG mediated to modern North Europeans by steppe).

    Both also seem low compared to formal stat based methods, so this method does not correct whatever discrepancy is there in ADMIXTURE compared to formal stat based methods. Lazaridis and Reich raised the issue that in supervised ADMIXTURE, it appears that error is present in which – “The estimation error (estimated – real ancestry) is strongly correlated (r=0.91) with the estimated Steppe ancestry, allowing us to predict it by a regression (Figure 1c), which indicates upward bias for high Steppe ancestry estimates and downward bias for low ones.” –

  6. Razib,

    Saying that the shared Steppe ancestry between Northern Europeans and South Asians is the factor that approximated them genetically is hard to understand?
    Saying that Dystruct idiotically building Yamnaya as Kostenki is hard to understand?
    Saying that, Europeans now have been on a trend to re-select their Mesolithic Hunter Gatherer genes instead of EEF and CHG ancestry that’s slowly being phased out in the same process that happened with Neanderthals and that this can be seen in both ADMIXTURE and PCAs is hard to understand?
    Saying that migrations from Anatolian Farmer and Natufian rich areas in the recent past to Europe (such as in Roman times) is a force that’s pulling those European populations more towards AF/N genetically is hard to understand?


  7. Saying that, Europeans now have been on a trend to re-select their Mesolithic Hunter Gatherer genes instead of EEF and CHG ancestry that’s slowly being phased out in the same process that happened with Neanderthals and that this can be seen in both ADMIXTURE and PCAs is hard to understand?

    yeah, i see what you are saying.

    1) you still sound dumb. sorry.

    2) this isn’t happening. i know this for a fact (if you are not being dumb and are talking background selection due to slight deleterious effect. you really don’t know how to say this correct even when you have a non-dumb idea).

    3) feel free never to leave a comment here again.

Comments are closed.