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

Open Thread – 11/16/2020 – Gene Expression

Posted on by Razib Khan

Tomorrow Rhythm of War is out. This is the fourth of Brandon Sanderson’s ten-book Stormlight Archive. I began reading this series ten years ago, when my life was different. I don’t know when I have time to read this, but I will.

Sanderson’s work is not dark, and I need that now. Is it a coincidence I became a fan of George R. R. Martin in the 1990s?

Apologize for getting behind in the book club. I’ll catch up!

The main reason is I’m kind of overwhelmed now, especially getting my Substack up (this week!). I plan to put long-form essays and particular podcasts there. And more comment engagement. Thanks to everyone who subscribed before the announcement! (I’ll blast out an email to the Mailchimp lists soon)

Matt Yglesias moved to Substack.

Another allocation of time: I’ll have a review of Alex Gibney’s new film, “Totally Under Control”, in Quillette soon. It’s OK. But kind of thin.

I’ll be taking suggestions on what to write about, but probably the two first essays will be a “short genetic history of the world” and the “closing of the liberal mind.”

Related Posts:

25 thoughts on “Open Thread – 11/16/2020 – Gene Expression”

  1. As for bearish on America, how does Canada look for graying Americans who are attached to North America’s vast and plenty? Asking for a friend.

  2. in a warmer world i think it’s a good bet. also, america has scale problems. too many ppl in a very democratic system without enuf intermediation. canada is big, but not so many ppl

    yet

  3. “I’ll have a review of Alex Gibney’s new film, “Totally Under Control”, in Quillette soon.”

    I RSS Quillette. I read it on Friday. It was dated November 13.

    Here was the link:

    https://quillette.com/2020/11/13/totally-under-control-a-review/

    I just tried the link and it is 401.

    I was going to congratulate you for being able to break the news gently that with COVID-19, as in Hollywood, nobody knows anything.

  5. Zombardment: Canada is TFC. It is also overrun with the woke lunatics who are such a plague in the US.

  6. Everybody’s going to Substack.

    I read the first book of Stormlight Archive, not going to risk it with the rest until BS finishes at least eight of the planned 10. (Invested into both WoT and ASoIaF, got burned).

  12. @Rob, what new information about the world did you get out of that one, if you don’t mind the question? Like, what did you learn from it, if anything?

  14. aha fair enough. Clicked on it and it made me salty enough to quickly stop watching, so was wondering if anything more meaty there (against expectations!).

    I guess with me (though nobody asked!) stuff like that seems a bit pointless to go over, as it’s like everyone knows Woke Is Real by now, and what they believe about it (good/terrible and crazy). Or else if they don’t think its a thing, or they just don’t think it’s important (“Meh, it’s merely a distraction from the intractable problems of Late Stage Capitalism!”etc.), then there’s no changing their minds at this stage!

    (And it probably wouldn’t matter if anyone else did change minds; I’m gonna guess that, like the time to do anything effect to prevent Covid was largely with travel bans and distancing before it entered, the time for advertising Woke’s madness and undercutting Woke “gateway beliefs” in ways that would change the course of future events probably like a generation ago now, in 2008-10 or smth. The cake is baked and sooner or later we will all eat our portions.).

    So videos like that just seem like an equivalent to spiralling reading over the worst depths of 4chan or something (with incongrously slick Woke-media production values). But other people might vary, not saying you shouldn’t post! (And I should probably to learn to say all the above in one paragraph!)

  16. @PD, yes, that paper’s result seems pretty interesting. It’ll need some more analysis to determine how much it any actually has much effect on IFR for this coronavirus.

    Would note, even this coronavirus should have relatively low effects on fitness, as predominantly hits old individuals, in particular individuals who would mostly (not all, but mostly!) be already deceased from another cause earlier in life in HG and pre-modern societies through history, and would at any rate be well beyond reproductive age (though not grandparenting age). So any sustained effect would have to be from nastier viruses or fairly small fitness advantage over a long time?

    It does raise plausibility that a number of nasty CV have been about in East Asia for a long time though, and that this may have been the case even recently.

    I do wonder how well it matches up with apparently really nasty fatality rates among Native Americans and Pacific Islanders though. Both of those groups have most of their ancestry from populations who were evolving in East Asia around 22.5 kya (about 70% in each case) when the paper presents the effect.

    To use the paper’s phrase “The peak around 900 generations ago (870 generations more exactly) spans approximately 200 generations, where the pink distribution is clearly above the blue one”… “900 generations (~25,000 years) ago and exhibit a coordinated adaptive response that lasted until around 200 generations (~5,000 years) ago”

    Most of the ancestry of Native Americans was coevolving in East Asia probably until 16kya, so I guess that is a relatively early branch off, but most of the Austronesian Pacific Islander ancestry would be coevolving in East Asia for the entire period! (Austronesian expansions did not get out of China until 5000 YBP).

    There could be some explanations for this – high obesity, general relaxation of selection, etc.

    Their paper seems to find no selected response in the Peruvian genomes in 1000G, which are mostly Native American. That seems somewhat strange if the effect really happened when they think it did. But it could be explained (little effect had happened before NA ancestors diverged). It would be good to repeat in a set of Polynesian genomes, which would provide a more interesting test (as later divergence point).

  17. They do note that “Accordingly, CLUES estimates very high selection coefficients in the interval between 1,000 and 500 generations ago” – 27k to 14k – “(Dai average s = 0.05, Beijing Han s = 0.11), but much selection coefficients from 500 generations ago up to the present (Dai average s = 0.003, 281 Beijing Chinese s = 0.004; Figure 4A, B).”.

  18. @Matt: I’m one of the less technically-savvy readers here, mostly lurking for unrecorded historical insights. I appreciate your thoughts.

    I do wonder if there is a basic model of (A) long-term genetic differences creating variability in susceptibility to coronaviruses and (B) relatively recent exposures to coronaviruses significantly aiding an immune response. The sweet spot would be for both (A) and (B), but some populations might have only one or none. There was a paper last week (below) suggesting that Sub-Saharan Africa has experienced relatively few infections due to prior exposure to coronaviruses, so I wonder if (B) might be relatively more important than (A):

    https://www.sciencedirect.com/science/article/pii/S1201971220323109

  20. Nice interview Robert.

    @PD, yes, that’s possible for some, and yep saw that Kenya paper and had some similar thoughts to your own as possibilities.

    Francois Balloux gave a thought on the paper, noting a result from the ONS that proposed that Chinese in Britain did not have lower infection and fatality rates than expected – https://twitter.com/BallouxFrancois/status/1329696239940677636 . The objection to this when I’ve raised it before with someone else (that time it was in response to a claim that East Asians had a behavioural advantage in infection) was that this was a small sample though. I think that’s fair and would need a robust re-examination in fatality data so far – it surely can’t be that hard to just infer or determine “Asian” sub-ethnicity and then check fatality rates vs expectation…

    (Determining real prevalence of infection:fatality may be hard due to lots of structure in infections, e.g. – https://www.medrxiv.org/content/10.1101/2020.11.12.20230763v1.full – ” 956 health care workers University Hospitals Birmingham NHS Foundation Trust, UK … Age, sex, number of household co-occupants, public transport use and index of multiple deprivation scores associated with participants home postcodes did not significantly influence seroprevalence (Table 1, Supplementary Figure 1). However, ethnicity did have an effect with individuals of Black (72.2% seropositive 95% CI 56.0 – 84.2%) and Asian ethnicity (54.1% seropositive, 95% CI 46.2 -61.4%) demonstrating the highest seroprevalence (overall Chi square 19.2, degrees of freedom 5, p=0.002) (Supplementary Figure 1).” (Likely South Asian hospital workers, because Birmingham, England). Huge rates, and must be related to type of employment?

  21. Matt, I have a question for you. The F-stats/fixation indices method is several decades old but when were f4 and f4/D-stats introduced?

  24. Putting it on record: https://en.wikipedia.org/wiki/Larry_Kwong

    The first POC in the NHL was an Asian guy but no one cares cuz he’s not black like Willie O’ree (who gets all the press.) The 2nd POC NHLer was First Nations so no one’s heard of him either:) Willie O’ree gets all the attention. The erasure of Asians continues!

    I think i’m gonna start taking Vitamin D again:
    https://old.reddit.com/r/science/comments/jxvack/analysis_of_vitamin_d_level_among_asymptomatic/

Comments are closed.