From Undark, a lengthy thumbsucker worrying that Genome-Wide Association Studies could be used Wrong, such as for predicting IQ and other interesting traits.

From a Fledgling Genetic Science, A Murky Market for Prediction

The commercial push is raising tension between scientists seeking new genetic links, and entrepreneurs seeking profit.

BY ASHLEY SMART
10.27.2023

Ashley Smart is the associate director of the Knight Science Journalism Program at MIT, and a senior editor at Undark.

The article profiles a start-up that advises customers on what the exploding literature on GWAS says about their genome (or whoever owns the genome they submitted—maybe they got it off a fork or glass?).

The article is concerned that people are interested in interesting topics like IQ:

IMPUTE.ME WAS NOT just a health site. Among the analyses the site offered were genetic scores relating to athletic ability, hypersexuality, intelligence, and other categories that geneticists and bioethicists have deemed of heightened concern due to their perceived ties to social status, their associations with harmful stereotypes, or their centrality to the identities of minoritized groups.

The intelligence scores Folkersen offered clandestinely—at a page hidden from the Impute.me homepage but accessible with a URL or direct link. Responding on Reddit after a user seemed to express disappointment at being ranked in the zeroth percentile for the trait, Folkersen explained his rationale: “This is exactly the reason why the intelligence module is unlisted,” he wrote. “People can’t handle it if they get a low score because they somehow take it more personally than a real IQ test. They are wrong.”

Folkersen argued that an IQ test is a far more accurate prediction of intelligence than a polygenic score. Indeed, many scientists have said that polygenic scores aren’t equipped to predict individual outcomes for traits like intelligence. The genetic variants identified in a large 2018 GWAS for intelligence, published in Nature Genetics, could account for around 5 percent of the variance observed in studied populations of European ancestry, and part of that may reflect factors that are only indirectly related to genetics.

Despite Folkersen’s efforts to keep them discreet, Impute.me’s intelligence scores garnered attention in internet forums and message boards, at times providing fodder for racist diatribes. Emil Kirkegaard—a right-wing blogger who has argued for innate intellectual differences between races—described the intelligence scores as the “juicy parts” of the site.

Of course, at present GWAS studies are full of noise and explain only a small percentage of total variance for many traits of interest. The sample sizes required to do GWAS on a powerful scale are, presumably, immense, in the tens of millions. (Or maybe we’ll never get all that much out of this technique.) So, currently, there isn’t all that much you can actionably do with GWAS data.

We know from twin and adoption studies, that genes are important. But these studies don’t open the black box to tell us which genes do what, while GWAS has made some progress in doing exactly that. But even sample sizes in 7 figures tend to leave an awful lot unexplained.

Would customers understand that their results might be a very small number that barely moves the needle in terms of estimating your odds of getting Alzheimer’s or whatever?

On the other hand, if you were a homosexual looking to hire a sperm or egg donor, something is better than nothing. Gays and lesbians are sacralized identities, so there may come a clash between LGB would-be parents and the growing movement to lock down much of genetic research for fear it might tell something about topics of general importance, such as IQ.

On the other other hand, you can tell a lot about genotype by looking at phenotype. If you want to know whether a potential egg donor has high IQ genes, give her an IQ test. You’ll learn a lot more from an IQ test than from a GWAS analysis at present.

But it’s not unreasonable for start-ups to have a business plan devoted to catering to hobbyists rather than to serious uses in its early years. People should have a right to learn about their own DNA. If they find interesting the small amounts that can currently be gleaned about many traits from GWAS, so be it. Maybe in the long run, this will prove highly useful, so it’s fine if hobbyists fund the long term right now.