“The genes can set the rules, but the outcome will vary ... This is especially true in the brain, due to the nonlinear, self-organizing nature of development, where small differences at one stage can have cascading consequences & be amplified across devo.”https://blogs.scientificamerican.com/observations/is-our-future-really-written-in-our-genes/ …
-
If you think that he overvalues, say, 10% of the variation explained, fine, but when he says that 10% variation is explained that is unambigiously opposed to the statement "DNA can predict your future from birth with 100 percent reliability."
The reliability refers to the lack of measurement error in polygenic scores I think. No other psychological measurement has as low measurement error as PGS.
-
-
What is the PGS measurement error? I don't think people report it, and it might not be calculable. What is the likely measurement error of taking a weighted sum of thousands to 100s of thousands of regression coefficients?
-
It's very small and usually not reported because studies only measure them once. However,
@razibkhan could calculate some for you because he did multiple array tests on himself. It will be close to the genotyping accuracy which is usually >0.99. -
I don't think you can see it as technical error alone. These scores measure association with a phenotype, not just genotypes. Genotypes are random variables, but so are each of the thousands of individual weights (the GWAS coefficients)
-
That's what Plomin meant I think. You're now talking about the estimation error in the PGS model. That's a different issue all together and Plomin did obviously not mean our current PGS models are perfect (capture 100% heritability).
-
Well, I'm not talking about prediction error. I think we are looking at PGS from different angles, or with definitions that are too fuzzy. Twitter ain't great for ironing out this sort of disagreement..
End of conversation
New conversation -
-
-
Yes, I believe so, too. But I think it very dishonest to put this notion of "reliable" into the same sentence as the accuracy of the _prediction_.pic.twitter.com/lsBo6HLIZg
-
Also, does Harden or Mitchell disagree about the lack of measurement error? I assume not, but then what is the reason to bring it up in that context? Other than to hope that people will misunderstand, and it will make Plomin look bad.
-
I think it's great that Mitchell puts a spotlight on stochastic developmental noise. And how that is relevant for genetic prediction. I just don't see the need to characterize Plomin's views in this manner (which is at best highly misleading).
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.