Statistical geneticist @EdinUniLBC, Psychology Department, The University of Edinburgh.
Edinburgh, Scotland
Joined November 2016
- Tweets1.7K
- Following227
- Followers873
- Likes1.4K
Pinned Tweet
Our new UK Biobank GWAS on Intelligence is out now! 187 loci associated loci! nature.com/articles/s41380-0… @uk_biobank @patrickaturley @rjhaier @_r_c_a @genscot
W David Hill retweeted
Out now in Nature: causal, within-family effects explain ~all of the association between indigenous American (IAM) ancestry and height and type-2 diabetes but ~none of the association with education.
nature.com/articles/s41586-0…
The main results haven't changed from the preprint (see thread below). This is an important paper, the first real progress on studying the genetic contribution to group differences for a long time. In the revised manuscript, we expanded our discussion on how to interpret these results.
The study design uses random, within-family variation in proportions of ancestry from European (EUR), IAM, and African sources to estimate the effect of inheriting EUR vs IAM ancestry. The random variation comes from differences in ancestry proportions between full-siblings, and between offspring and the average of their parents' ancestry proportions.
This diagram shows the ancestries of the parents (Ap(i), Am(i)), two full-sibling offspring (Ai1, Ai2), the sibling phenotypes (Yi1,Yi2).
It shows that sibling ancestries are conditionally independent of the environment given the parental ancestries (implying that their difference is also independent). It also shows that the within-family ancestry effect of Aij on Yij can be mediated through Sij, the ancestry induced social environment of the sibling ij (c.f. the 'red hair gene' thought experiment of ancestry related social discrimination etc.), or it can 'directly' affect the phenotype, i.e. through other mechanisms, e.g. more strictly biological ones.
As we state in the discussion: "While our within-family ancestry effect estimates can be given a causal interpretation, they do not imply a specific mechanism. It is possible that ancestry differences cause phenotype differences through social mechanisms such as ancestry-related appearance-based discrimination.
...
Our estimates are consistent with ancestry associated phenotype differences in height and T2D risk being entirely explained by within-family ancestry effects, but the absence of a between-family effect does not imply or prove a mechanism by itself. When we fitted educational attainment in the model as a proxy for social environment, there was no change in the estimates of the within-family effects for height and T2D, implying that any hypothetical social mechanism is not correlated with educational attainment."
Personally, I find the lack of a between-family ancestry effect for height and T2D quite convincing that the mechanism is not primarily the ancestry induced social environment. For this to be true, one would have to believe that within-family variation in ancestry induces social environment effects with large effects on height and T2D, but that the ancestry of one's parents — and the social environment induced — is totally irrelevant.
Nonetheless, this method can be applied to more controversial questions, and establishing the mechanism through which within-family ancestry differences cause phenotype differences will be necessary before any controversies can be settled.
W David Hill retweeted
Excited to share our work on genetic nurture, where we find that parental genetic liabilities associated with educational attainment appear unlikely to exert a large genetic nurture effect on childhood cognitive performance.
Preprint: biorxiv.org/content/10.64898…
To reach this conclusion, we addressed a statistical problem in trio-based polygenic score regressions widely used in human genetics. In these, one regresses the child’s outcome on the child’s, mother’s, and father’s polygenic scores, and associations with parental PGSs are often interpreted as suggestive evidence that parental genotypes influence the child through the environment (genetic nurture).
However, assortative mating on the PGS trait can induce “phantom genetic nurture” in this design, as previously described by @AlexTISYoung and others. Intuitively, this occurs because PGSs are noisy and incomplete, leaving much of the genetic liability they target unobserved. One parental PGS can ‘tag’ the other’s genetic liability, which in turn tags some of the child’s genetic liability not captured by the PGS. Thus, a parental PGS association in the trio regression could simply reflect the child’s genetic liability influencing their trait.
To address this, we developed RAVEL (Recovering Assortment-adjusted Variance Explained by genetic Liabilities). RAVEL uses two independently constructed PGSs for the same trait. Both noisily index the same underlying genetic liability, but their estimation errors are independent. Under explicit assumptions, this design allows us to recover assortment-adjusted associations of the child’s and parents’ latent genetic liabilities with the child’s outcome. In simulations, RAVEL recovers the target effects well across the settings we examine.
For the empirical analysis, we used >30,000 complete trios in the MoBa cohort to assess whether parental genetic liabilities for educational attainment (EA) show evidence of genetic nurture effects on childhood cognitive performance. With standard trio PGS regression, we find, as others have, that the parental PGSs together explain ~10% as much variance as the child’s PGS (excluding covariances).
However, after applying RAVEL, the maternal association attenuates completely, but a small, significant paternal association remained, with the parental liabilities inferred to explain less than 0.5% as much variance as the child’s.
In short, the standard analysis suggests a potentially meaningful genetic nurture effect, whereas the assortment-adjusted analysis suggests that any remaining EA-related nurture effect on childhood cognitive performance is likely very small.
Much more in the paper, including evidence for a possible maternal effect on preterm birth, sibling IBD regressions, and differential parental associations at the phenotype tails. We are significantly expanding the analyses before journal submission, so feedback is welcome! I am extremely grateful to Olivia, Laura, Alexandra, @hfsunde, and @hilsomartin for their contributions to this work.
W David Hill retweeted
'Geneticist' is an insult now is it? Insane.
W David Hill retweeted
Our review summarizing the field of the genetics of the human proteome is now published in @NatureRevGenet I would suggest pay attention to Figure 3b nature.com/articles/s41576-0…
W David Hill retweeted
See below for a very impressive preprint that tries to summary the genetic basis of healthcare costs (to which I made a small contribution).
W David Hill retweeted
SV-GWAS is highly accessible now! Our new paper in @NatureGenet shows how HiFi long-read assemblies let us repurpose SNP-based GWAS data to impute structural variants (SVs) to interrogate their role in human complex traits and diseases. @WeiyangBai nature.com/articles/s41588-0…
W David Hill retweeted
Very excited about this preprint that we just posted! It introduces the Genomic-Relatedness Matched Association (GRMA) study. It’s an extension to family-based GWAS that uses extended relatives beyond only siblings in diverse-ancestry data with very little bias.
W David Hill retweeted
Just published, new paper on using polygenic intelligence scores from adults to chart cognitive development in children: icajournal.scholasticahq.com
W David Hill retweeted
Important thread about real world utility of polygenic embryo selection.
W David Hill retweeted
This conversation is highly recommended, especially if you have heard of Charles Murray but never read any of his books.
🚨 New podcast episode! 🚨
Charles Murray (The Bell Curve) doesn't shy away from the hard questions, and neither do we. Topics include:
✅ IQ, genetics, & human differences
✅ The Truth behind The Bell Curve
✅ Faith, meaning, & Murray's personal journey
youtu.be/lqOhOvw0kbs?si=5hO3…
W David Hill retweeted
Family-GWAS results were critical in testing the robustness of inferences of natural selection in the recent Akbari/Reich paper.
We're releasing family-GWAS sumstats on 28 phenotypes on thessgac.com/ data portal, using the powerful method we developed in Guan et al.
W David Hill retweeted
Mr. Gladwell never seems to get it right.
Lewis Terman can't defend himself against this travesty from Malcom Gladwell, but we can do it for him.
Here's the real study results from Terman's own volumes (Genetic Studies of Genius) and later analyses.
Terman identified ~1,528 California schoolchildren with Stanford-Binet IQs of 135+ (average around 150).
He followed them studiously for decades with follow-ups continuing after his death.
His goal was to replace myths about "geniuses" with data on their development. People thought IQ might cause lower health, physical weakness, susceptibility to disorder. Terman tracked not only IQ but personality, health, family background, and life outcomes.
The result? The group as a whole crushed general population benchmarks.
By their mid-30s (and this is close to the Great Depression) ~70% of the men and ~67% of women had bachelor's degrees vs. ~8% nationally at the time
Close to a 10x out performance !!
At a time when few entered let alone stayed on at University, many "Termites" pursued and achieved graduate degrees (97 PhDs, 57 MDs, etc.).
By their mid-40s, 96%+ of men were professionals or semi-professionals (doctors, lawyers, engineers, academics, etc.).
Incomes were roughly double the national white-collar median. Imagine: A single level to pull and you double national incomes!
They produced thousands of scientific articles, books, and patents. This was then replicated in wonderfully detailed and prodigious Vanderbilt studies.
Their health, far from poorly, was better and they had lower rates of divorce and fewer psychiatric problems.
The sample, which based on Gladwell's presentation we should expect from regression to the mean to be downwardly socially mobile , was massively upwardly mobile.
Even relative to their (already fine) childhood homes. High IQ predicted climbing the socioeconomic ladder, not coasting on family money.
Hoping to obfuscate all this success, Gladwell spins a yarn for y'all.
His claim that "rich smarts" had a silver spoon in their mouths while poor smarts were "utter failures" (imagine saying THAT in 2026…) is nonsense and betrays Terman's own words and findings.
Terman did examine differences in adult achievement (A: highest ~top third or so; B: middle/moderate professionals; C: lowest ~bottom 20–30%).
There was a correlation with childhood family socioeconomic status (SES): A's had more parental education/support/resources on average; C's had less.
But the "C" group ("failures" according to Gladwell) far from "produced nothing" Most won college degrees. Most won professional or semi-professional jobs. Most had incomes and achievements well above the general population. Framing these people as underachievers is a disgrace, TBH.
Terman was one of those leading the emphasis on both the need for opportunity and a society which encouraged education, had capital to allocate, enforced the rule of law etc.. And to emphasize the role of "non-cognitive" traits of conscientiousness (persistence, drive, goal setting and ambition). He also emphasised the need for good health.
He didn't pretend that we do not stand on the shoulders's of giants (those who gifted us our current SES) but he did show in staggering detail, the amazing accomplishments of all of these children, identified at a young age as having already college-level knowledge and ability.
Puff like poverty "reduces a one-in-a-billion brain to a lifetime of worse than mediocrity" is debunked in the American context by the whole cohort outperforming expectations massively. Not just the tippy top, but the cohort as a whole . Resoundingly.
Terman, and no scientist you will find, ever claimed IQ alone guarantees universal leadership or output - the idea of "just one thing", here as everywhere is a distracting red herring.
What Terman did was revolutionize what we know about just how strong a predictor of life success high IQ is. And how precious those point-1 percent are.
It is curious that Terman missed William Shockley and Luis Alvarez (future Nobel physicists). Shockley was co-inventor of the transistor and father of Silicon Valley (along with Frederick Terman, the son of Lewis Terman, who in turn created the "Stanford Binet" IQ-test and did much to promote Stanford and IQ).
As the late Danny Kahneman would have point out, calling out this miss of two people in a sample of 250,000 is to commit the fallacy of low base-rates: Sampling ~1,500 out of ~ 250,000 kids makes missing a 2 people in the extreme tails a likelihood, not a fatal flaw.
This is classic Gladwell: Storytelling with punchy examples, but under scrutiny, the specifics on Terman all fall apart.
Unlike Gladwell, Terman's results stand the test of time far better than any pop-psych retellings.
h/t @charlesmurray
W David Hill retweeted
New from David's Reich's lab: "In the past ten millennia [in Europe], we find that many hundreds of alleles have been affected by strong directional selection," including those influencing human intelligence.
nature.com/articles/s41586-0…
W David Hill retweeted
Today the worlds most powerful genetic predictor of IQ, CogPGT, has been published in the peer reviewed journal Intelligence and Cognitive Abilities.
When used for embryo screening, it can substantially boost expected IQ of future offspring.
Read on for the scientific details!
W David Hill retweeted
i mean, this should be pretty self-explanatory but still a good read.
(i.e., as traits become more polygenic, the relative contribution of localized, gene-proximal (especially exonic) effects decreases, while heritability becomes increasingly dominated by distributed, largely noncoding regulatory effects across the genome, with substantial and relatively stable contributions from intronic regions.)
W David Hill retweeted
StratGWAS is our new tool for more efficient GWAS of heterogeneous diseases. Instead of treating cases equal, it weights them based on relevant phenotypic information such as medication use, age of onset or recruitment strategy. Full details on MedrXIv medrxiv.org/content/10.64898….
W David Hill retweeted
Here's the longer version of our Nature piece.
Our argument is simple: statistical approximation is not the same thing as intelligence.
Strong benchmark scores often say very little about how LLMs behave under novelty, uncertainty, or shifting goals.
Even more importantly, similar behaviors can arise from fundamentally different processes. In another paper, we identified seven epistemological fault lines between humans and LLMs.
For example, LLMs have no internal representation of what is true. They often generate confident contradictions, especially in longer interactions, because they do not track what is actually true.
Another example. Yes, LLMs have solved some open mathematical problems, but these cases typically involve applying known methods to well-defined problems. LLMs cannot invent anything that is truly new and true at the same time, because they lack the epistemic machinery to determine what is true.
None of this means LLMs are useless. Quite the opposite: they are extraordinarily useful.
But we should be careful about what they are and what they are not.
Producing plausible text is not the same as understanding.
Statistical prediction is not the same as intelligence.
So despite the hype from the usual suspects, AGI has not been achieved.
*
paper in the first reply
Joint with @Walter4C and @GaryMarcus
W David Hill retweeted
🧵 New publication from the PGC Anxiety Working Group. Our GWAS meta-analysis of anxiety disorders is now published in @NatureGenet! 🔗: doi.org/10.1038/s41588-025-0…
W David Hill retweeted
Excited to share our preprint introducing COXMM! COXMM is a Cox proportional hazard mixed model for estimating the heritability of time-to-event (TTE)/longitudinal traits. 🧵1/10
Work with @SashaGusevPosts and @sr_sankararaman.
doi.org/10.64898/2026.02.16.…
W David Hill retweeted
Pretty nice obituary for James Watson
sciencedirect.com/science/ar…
The most significant scientific discovery of that time — perhaps any time in the life sciences — was his and Francis Crick’s: the double helical structure of DNA, and the base pairing it contained, the feature that answered the two great mysteries of life — how genetic material can both copy itself and also carry the code that defines the characteristics of living organisms.
…arguably the most influential scientist of the second half of the 20th century, his impact felt through his scientific discoveries, his game-changing books, and his rare talent for institution building and administration.