"There are two exciting pieces of news in embryo selection. The technology is finally moving forward at a steady clip. Maybe we aren’t doomed by dysgenics after all!"
That is good news. Long-term, I am hopeful that genetic enhancement will advance humanity.
"This may be because we’re dealing with a children’s sample, or because autism has become “more popular” in recent generations, making it less elite. Who knows."
I believe that it is most likely that the spectrum for autism has been broadened over the years.
Good question. So using the Herasight tool at https://www.herasight.com/#embryo-tool, they show that with 10 embryos, one could with their technology expect about 7 IQ by picking the best out of 10 (assuming it implants 100%).
Re-reading Gwern's calculations at https://gwern.net/embryo-selection#polygenic-scores-for-iq, you see that these concern not the practicable gains but the theoretical optimal given some GCTA-variance. He found 33% in his meta-analysis, which would be higher if corrected for measurement error. As far as I can tell, assuming PGS with 33% variance explained and 10 embryos, he finds an expected gain of 9.4 IQ. This is higher than the ~7 IQ given by Herasight's tool. However, since their tool is based on whatever validity they put into it, maybe beta = 0.44, 19.4% variance, their gains should be smaller. I guess we could download Gwern's code and run it for all the values of his variance parameter to and map this to the PGS beta to see visually how PGS gains affect expected IQ gains at a given set of embryos, and whether the value is consistent with Herasight's calculator.
"There are two exciting pieces of news in embryo selection. The technology is finally moving forward at a steady clip. Maybe we aren’t doomed by dysgenics after all!"
That is good news. Long-term, I am hopeful that genetic enhancement will advance humanity.
"This may be because we’re dealing with a children’s sample, or because autism has become “more popular” in recent generations, making it less elite. Who knows."
I believe that it is most likely that the spectrum for autism has been broadened over the years.
Thanks for the interesting article.
Isn’t the 9 point advantage from 10 embryos roughly what Gwern estimated was possible with much older data? What’s the game changer in this story?
Good question. So using the Herasight tool at https://www.herasight.com/#embryo-tool, they show that with 10 embryos, one could with their technology expect about 7 IQ by picking the best out of 10 (assuming it implants 100%).
Re-reading Gwern's calculations at https://gwern.net/embryo-selection#polygenic-scores-for-iq, you see that these concern not the practicable gains but the theoretical optimal given some GCTA-variance. He found 33% in his meta-analysis, which would be higher if corrected for measurement error. As far as I can tell, assuming PGS with 33% variance explained and 10 embryos, he finds an expected gain of 9.4 IQ. This is higher than the ~7 IQ given by Herasight's tool. However, since their tool is based on whatever validity they put into it, maybe beta = 0.44, 19.4% variance, their gains should be smaller. I guess we could download Gwern's code and run it for all the values of his variance parameter to and map this to the PGS beta to see visually how PGS gains affect expected IQ gains at a given set of embryos, and whether the value is consistent with Herasight's calculator.