There are at least two studies I’m aware of that are attempting to untangle the genetics of high IQ. They both involve looking at mathematical and sciencey types of people. I don’t think they’ll be that successful*. One reason has to do with covariance. But another reason is that I think these studies might be looking at the wrong type of people. Usually, mathematical aptitude is thought to be related to fluid intelligence–solving novel problems, identifying patterns and relationships, and applying deductive and inductive logic (though the JUMP program’s success at improving mathematical ability would argue this is absolutely not the case). Crystallized intelligence, on the other hand, involves memory, skills, and experience. Vocabulary and general knowledge are good proxies for crystallized intelligence.
What’s interesting is that a recent study argues that tests typically thought to focus on fluid intelligence show very little heritability, while tests that emphasize crystallized intelligence have a higher heritability. This makes sense when we consider the Flynn Effect–the three points per decade IQ increase that has occurred during the last century (and, yes, that means an average twenty year old in 2010 (IQ = 100) dropped into the midst of twenty year-olds from the Greatest Generation would rank at the ninth percentile–IQ = 121. Your grandparents are right; you’re smarter than they are). What’s notable about the Flynn Effect is that it’s mostly an increase in fluid intelligence, not crystallized intelligence.
So I would argue if we stand any chance of finding a genetic basis, we need to focus on things like a strong vocabulary and linguistic skills, as well as memory and recall.
Forget the scientists, get some poets (and probably actors too).
*Both studies are ludicrously underpowered, meaning if anything is found, it’s probably a statistical fluke.