tag:blogger.com,1999:blog-227780861638767023.post89906781858704943..comments2024-01-27T00:27:45.851+00:00Comments on race/history/evolution notes: The Atlantic: How Eugenic Breeding Transformed the Dairy Industryn/ahttp://www.blogger.com/profile/02378473351485233448noreply@blogger.comBlogger6125tag:blogger.com,1999:blog-227780861638767023.post-87423572347818858532012-08-21T13:40:32.196+01:002012-08-21T13:40:32.196+01:00If you look at table 2 of the IQ paper, he predict...If you look at table 2 of the IQ paper, he predicts IQ phenotypes for a subsample left out of the prediction model with a correlation to the true value of around 0.1 for fluid intelligence. <br /><br />His method of estimating heritability is based on the animal breeding model, which fits a linear mixed model to all the variants simultaneously. This allows one to construct a best-linear-unbiased-predictor from the fittted SNP effects, even if each SNP effect itself is not found to be significantly different from zero. <br /><br />It is certainly not accurate yet, but it has only been tried in a small GWAS sample. If you had a large sample with sequence data (or at least sequence imputed data), then you would be able to capture all of the genetic variability influencing intelligence thereby allowing for an accurate prediction of phenotype from genotype for a highly heritable trait like IQ.<br /><br />Re Silver, I'm sure it'll happen eventually, but probably not before the Chinese have bred a race of super-geniuses. Surely there's going to be strong challenges to any genetic selection method for IQ in humans in the west?Lemniscatenoreply@blogger.comtag:blogger.com,1999:blog-227780861638767023.post-76220676822662405012012-08-21T08:49:41.629+01:002012-08-21T08:49:41.629+01:00Of course, such a beneficial, voluntary thing will...<i>Of course, such a beneficial, voluntary thing will not be allowed to happen.</i><br /><br />Your pessimism is unwarranted (but apparently a staple postscript when it comes to eugenicist genes-talk). <a href="http://www.prnewswire.com/news-releases/beautifulpeoplecom-launches-virtual-sperm-bank-96783039.html" rel="nofollow">This gang</a> is going straight for the jugular: looks. (Look are for many an even more frightening prospect than intelligence.) Equalitarian reactionaries will not be able to hold back the tide of eugenic progress indefinitely, even if they did manage to win some impressive victories in the interim.<br /><br /><br /><br />Silvernoreply@blogger.comtag:blogger.com,1999:blog-227780861638767023.post-22935613172593593012012-08-21T03:19:57.270+01:002012-08-21T03:19:57.270+01:00That just calculates heritabiity. He can't pre...That just calculates heritabiity. He can't predict IQ from just the genotype very well. No one can yet.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-227780861638767023.post-75364661186230977952012-08-21T00:14:02.108+01:002012-08-21T00:14:02.108+01:00Here's the Visscher paper on intelligence: htt...Here's the Visscher paper on intelligence: http://www.lscp.net/persons/ramus/fr/GDP1/papers/Davies11.pdfAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-227780861638767023.post-43320140593646672132012-08-20T16:30:02.989+01:002012-08-20T16:30:02.989+01:00"Visscher's group has already used his &#..."Visscher's group has already used his 'realized relatedness matrix' to predict IQ in humans with non-negligible accuracy."<br /><br />Can you give a reference for this? Calculation of heritability is not the same thing as phenotype prediction from genotype.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-227780861638767023.post-59884224920830736042012-08-20T12:41:29.620+01:002012-08-20T12:41:29.620+01:00I've been working on Visscher's method, am...I've been working on Visscher's method, among others, for heritability estimation.<br /><br />It really is just a matter of sample size. Visscher's group has already used his 'realized relatedness matrix' to predict IQ in humans with non-negligible accuracy. This is using GWAS data in a relatively small sample, and GWAS data necessarily only captures a small subset of genetic variation, especially at rare variants. When we have large samples of sequence data with IQ phenotypes, it will be easy to predict someone's IQ 'breeding value'. <br /><br />If we had a free market in human sperm banks, I bet it wouldn't be long before they started advertising prime human sperm based on its breeding value for traits such as IQ, height, longevity, etc.<br /><br />Of course, such a beneficial, voluntary thing will not be allowed to happen.Lemniscatenoreply@blogger.com