2 ms·
> Given the number of genes involved in intelligence I would be more surprised if they didn’t share an underlying genetic basis. That doesn't follow. Simply b
by gwern 11y ago
> Given the number of genes involved in intelligence I would be more surprised if they didn’t share an underlying genetic basis.
That doesn't follow. Simply because the genes sets overlap doesn't mean they have to correlate. If there are 10k genes affecting schizophrenia and 10k intelligence (plausible scales from the GWASes), then since there aren't that many varying genes there is likely overlap, but there's no reason to expect simply from this overlap that they will covary; the genes conferring benefits for one could confer problems for the other. In fact, the overlap could be total without creating correlation; and it would be just as plausible to argue for anti-correlation/inverse-correlation - the same gene could protect against schizophrenia and reductions in intelligence by making neurons more robust or defending against infections. There should be some deeper reason for any correlation of IQ/schizophrenia than simply 'they both are affected by a lot of genes'. (Well, an awful lot of traits are affected by a lot of genes, you know... Polygenicity is expected to be the norm for most of the heritable stuff.)
Worse, there are GWASes for both schizophrenia and IQ; the former has made a lot more progress than the latter, but nevertheless, I am not aware that any particular overlap in hits has been noticed.