DNA tests can predict intelligence

westernpromise

https://www.telegraph.co.uk/science/2018/03/12/dna-tests-can-predict-intelligence-scientists-show-first-time/
Oh dear! He's not going to be very popular!

"Our study identified a large number of genes linked to intelligence. We were also able to predict intelligence in another group using only their DNA."

Clearly, fairness demands that unborn children who will be unfairly intelligent should be deprived of oxygen in the womb so that they are as thick and morally incompetent as the average Momentum-supporting envy monkey.

Everyone must be levelled down to be equally stupid, ugly and miserable. Envy will end only when there is nobody left to envy.

Zero_Gravitas

As usual, Kurt Vonnegut was way ahead of everybody...

https://en.m.wikipedia.org/wiki/Harrison_Bergeron

westernpromise

Zero_Gravitas wrote: »

As usual, Kurt Vonnegut was way ahead of everybody...

https://en.m.wikipedia.org/wiki/Harrison_Bergeron

Good article about the left linked at the bottom:
https://en.m.wikipedia.org/wiki/Crab_mentality

kinger101

I suspect the whoever wrote the Telegraph headline "...for the first time" didn't have the intelligence to do a PubMed search.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3182557/

John-K_3

kinger101 wrote: »

I suspect the whoever wrote the Telegraph headline "...for the first time" didn't have the intelligence to do a PubMed search.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3182557/

That does not (admittedly after only a quick skim) invalidate the headline. We know that intelligence is hereditary, but that is not the same as saying that looking at someone!!!8217;s DNA lets you predict it.

It is easy to see why; we could have no idea about which genes or combination of them was responsible for it, so be left flummoxed about how to predict it from the DNA (as opposed to the parents!!!8217; intelligence) despite knowing full well that it was heritable.

kinger101

John-K wrote: »

That does not (admittedly after only a quick skim) invalidate the headline. We know that intelligence is hereditary, but that is not the same as saying that looking at someone!!!8217;s DNA lets you predict it.

It is easy to see why; we could have no idea about which genes or combination of them was responsible for it, so be left flummoxed about how to predict it from the DNA (as opposed to the parents!!!8217; intelligence) despite knowing full well that it was heritable.

You're mistaken, although I can see the point you're trying to make. We've come a long way since Mendel's studies on peas.

Both studies were done using a technique called GWAS (genome-wide association study). It involves analysis the genomes of a large cohort to detect regions of the genome (loci) with a statistical association to a particular trait (in this case, intelligence). The number of DNA markers used in these studies typically exceeds the number of known genes in the genome.

By default, if you then pick any individual from within or outside of that cohort, and score all or only those loci related to the trait, the information you obtain is predictive for their intelligence.

The technique is so powerful that it's very likely to identify the actual genes underlying this trait, or if not, a close physical variation within the genome which is a reliable proxy. You don't actually need to know the gene to use it in a test. You just need a linked DNA marker.

I should add that the authors of the 2018 study are not making a claim of priority for the predictive ability of their study (that was added by the Telegraph journalist), whereas, the authors of the 2011 study do make such a claim.

It is the first to show biologically and unequivocally that human intelligence is highly polygenic and that purely genetic (SNP) information can be used to predict intelligence.

John-K_3

I!!!8217;m still not seeing it from that study. It is measuring heritability, but stops short of assigning enough significance to a prediction of an individual!!!8217;s intelligence based on their genome. The effect size is too small to make that claim.

!!!8220;Our findings imply that very large sample sizes will be needed to detect individual loci with genome-wide significance and that the majority of additive genetic variation for human intelligence is not explained by rare variants that are not in LD with common SNPs.!!!8221;

Again, though, my background is particle physics, not cognition and heredity, so I may still be misreading.

For non-scientists reading this, I see it as akin to saying that there are clear differences in height at the population level between males and females but that is not enough to predict a person!!!8217;s height based on their sex.

You can get an excellent measure of the relationship between height and sex without being able to make a prediction for one person.

Predictive for intelligence is (in my opinion) not the same claim as saying that you can actually predict a person!!!8217;s intelligence from it. That claim is a step further, as it claims that the effect size dwarfs other factors.

That said, I doubt that the new study crosses that line either.

LHW99

.................. define intelligence? ............

kinger101

John-K wrote: »

Again, though, my background is particle physics, not cognition and heredity, so I may still be misreading.

You are misreading;

Individuals with a higher predicted score had, on average, a higher phenotype. Thus, SNP effects estimated in the discovery cohort are significantly predictive of cognitive phenotype outcomes in a fully independent cohort.

Predictive for intelligence is (in my opinion) not the same claim as saying that you can actually predict a person!!!8217;s intelligence from it.

.

You'd never be able to get an absolute accurate predictor of intelligence because of the noise caused by environmental factors. But without any prior information, a DNA test would be better at ranking the intelligence of 1000 individuals picked at random, than flipping a coin. I appreciate the effects of the first test are small, but the p values were 0.028 and 0.092.

NB - there are other papers with QTL for intelligence, cited in the 2018 study as well. I only picked one example.

Arklight

LHW99 wrote: »

.................. define intelligence? ............

The ability to complete a paper based test in an exam room that a psychologist then marks and assigns a score, mostly.

The Right on here are very excited about IQ because a) they tend not to accept the limits of psychometrics because right wing people view the world in very black and white, and fixed ways; and b) they all assume that they would get massive scores if they ever did a test, which they won!!!8217;t.

Malthusian

LHW99 wrote: »

.................. define intelligence? ............

The ability to process information and derive a correct solution to a problem.