Straight talk about epigenetics | Hacker News Reader

Straight talk about epigenetics

razib.substack.com

65 pointsexolymph28 comments

I'm not qualified to give much of an opinion on epigenetics, but I do find it interesting how it is talked about in popular culture, and I think it's good for us to be careful about that.

At my last employer, we had a diversity training where the speaker framed the lesson around the idea of epigenetics and how we all inherited generational trauma. My BS-detector instantly went off, and I found it hard to take the rest of the session seriously. This seems to be a concept, like "quantum mechanics," that can be easily misused in general discourse.

> inherited generational trauma

It's becoming a favourite concept for all kind of bullshitters, you can't prove it and you can't disprove it, it gives you free card to claim what ever you like

The difference between the folk understanding of epigenetics and the scientic understanding of epigenetics is the timescale. To scientists, epigenetics is the process by which genes become 'permanently' silenced via modifications to the DNA like methylation, or the histones the DNA is wrapped around. This kind of long term silencing is clearly important if you want to have a body made of different cell 'types' - a muscle cell should not make the same proteins as a neuron, so you'd want to shut the neuron genes off, forever, in a muscle cell, and shut the muscle genes off, forever, in a neuron. Now most of these long term marks are actually removed when making germ cells. Sperm and eggs, at some point, can shut off enzymes like DNMT1 that maintain these marks, so that the next person can start from a fresh slate...

Now the FOLK understand of epigenetics seems to have glommed on to an actually very small part of the actual phenomenon: there are a small set of genes that can be methylated in these germ cells, and the silencing will be passed on from parent to child so that the child will only express a certain gene from the mother or from father. This is called 'imprinting'. The folk understanding then extrapolates this, to genes being marked for several generations. I think evidence for this is poor, and not really the focus of most of the 'concept' of epigenetics. But there is still research to be done.

Not folk, not imprinting.

Here you go: https://www.nytimes.com/2018/01/31/science/dutch-famine-gene...

Very long-winded. Sounds like "hard" epigenetics doesn't exist in humans (changes to DNA expression aren't literally handed down) based on that factory reset explanation, but "soft" epigenetics (hormonal signalling in utero leading to changes in gene expression) is very much on the table.

I believe you’re using “soft”v. “hard” to distinguish what the author calls “inter generational” v “trans generational”. Using this distinction, “soft” epigenetics in your example is just epigenetics. The fetus responding to an external stressor which happens to be induced by a stressor on the mother. And in the case of female fetuses who have a fixed number of eggs for life, can impact those and therefore one more generation.

“Deeper digging reliability shows that cases where epigenetic marks seem to have been inherited transgenerationally actually turn out to be conditional on the existence of a conventional DNA mutation being passed on within the family…” yet he doesn’t provide an example.

I found the book: Evolution in Four Dimensions: Genetic, Epigenetic, Behavioral, and Symbolic Variation in the History of Life by Eva Jablonka and Marion J. Lamb a great scientific introduction on how epigenetics works along with other methods of evolution.

Intergenerational as referred to by Trudeau doesn’t mean genetic or epigenetic. The grandchildren of people drafted for the Vietnam War show impaired earnings compared to a control group, and of course the descendants of slaves still suffer. Tear likely has to do with reduced inherited wealth, social capital and opportunities, not (epi)genetic trauma, combined with the dark legacy of Jim Crow, redlining and other discrimination.

> Diagrams or processes that apply to hundreds of distinct products could be cross-referenced and only appear once

First, this is an interesting metaphor. (Some other comments here say the article is “long-winded”. Saying this misses a key goal of the author; namely, to use a detailed metaphor to make the topic more approachable.)

Second, does research have good estimates on how much the human genome _actually_ reuses parts of itself across different biological processes?

"That doesn’t mean that transgenerational epigenetic transmission doesn’t happen"

This whole article was a long winded way for the writer to look like a smarty pants but in the end he could have just said; "We don't know, but transgenerational epigenetic transmission is possible.

No, he’s saying “we can’t rule it out, but the evidence doesn’t currently support the assertions being made about trans generational epigenetic transmission.”

consider also that nurture is a major confounding factor.

I am skeptical in part because I should think that if the Soviets could have demonstrated Lamarckism, they would've.

> if the Soviets could have demonstrated Lamarckism, they would've.

They tried to demonstrate a form of Lamarckism - Lysenkoism, resulting in countless deaths across Siberia, especially among ethnic minorities that were deported there.

https://www.encyclopedia.com/science/encyclopedias-almanacs-...

The title of his article includes "You can’t take it with you". That is not anything close to implying "we can’t rule it out".

I know the evidence does not totally support epigenetic transmission, but it is pointing towards that conclusion. What we need is more evidence, not dismissing it on the basis of not knowing.

Besides, it just makes sense biologically. Why would we not have a biological system that prepared offspring for risks or even changes in the environment?

> I know the evidence does not totally support epigenetic transmission, but it is pointing towards that conclusion.

The point of the article is the opposite: that the evidence is pointing away from epigenetic transmission. Acknowledging that we can’t yet rule out epigenetic transmission doesn’t undermine the point that the existing evidence is pointing away from it.

What evidence in particular? I scanned the article but couldn’t find much.

> Besides, it just makes sense biologically.

I think the author is trying to pursuade us that this sort of instinct is not sufficient to base large social theories on.

The comment above unnecessarily casts the author in a negative light:

> This whole article was a long winded way for the writer to look like a smarty pants …

There are other, simpler explanations; for example, the author might have wanted to write a longer article with detailed metaphors to make the content more approachable. It is clear the author wants to push back on pseudoscientific interpretations of epigenetics. I commend the author on this effort.

Criticism (even if perceived as ‘negative’ by some) is important; still, I’d ask that commenters dial back their sense of _certainty_ somewhat. Before assuming some flaw of the author, at least consider that (a) _you_ (and all your preferences about length and style) may not be their target audience and (b) the world of writing has many driving factors —- not all of which involve an author trying to “look like a smarty pants”

A much better lay-accessible introduction to epigenetics would be Ewan Birney's tweetorial from earlier this year [1] or his Guardian column from a few years ago [2].

Rabiz Khan lacks formal training in medicine or genetics, and I perceive him to be a crank who has a weird fixation on Jewish genetics and a history of writing racial superiority op-eds for the alt right community.

[1] https://twitter.com/ewanbirney/status/1520414171334168577

[2] https://www.theguardian.com/science/blog/2015/sep/11/why-im-...

Razib Khan was a PhD candidate in genetics at UC Davis - https://www.linkedin.com/mwlite/in/razibkhan , he’s pretty qualified to speak on genetics IMO.

He has no first-author publications that aren't op-eds [1], and no publications whatsoever from his time at UC Davis. Since dropping out afaict he's done no relevant academic or industry work. I'm skeptical of his qualifications over any other bio undergrad.

fwiw, I don't think credentialism is the strongest signal here. The larger disqualifying factor is his history of crank takes and writing for anti-semitic and alt-right blogs.

[1] https://scholar.google.com/citations?user=6Ye06ekAAAAJ&hl=en...

This is all what's known as "ad hominem". Discredit the argument, not the person.

what about the rat/cherry blossom study? Widely reported 10 years ago, seemed compelling ... did it find later support? skepticism?

https://www.newscientist.com/article/dn24677-fear-of-a-smell...

I didn't even know this is a thing. But considering the traumata that humans have survived and the extreme genetic bottlenecks, it s fair to say that nobody has the right to claim they are more traumatized than others just because there is written history of their ancestors

(Am I to understand by the downvotes that people actually make such claims?)

Reply on news.ycombinator.com