Max Castle

I know you pretend to be genetics literate on the internet, but that paper doesn't show what you think it shows. Try again.

Mind explaining what it does show for those of us who can't even pretend?

These are GWA (genome-wide association) studies that essentially correlate genetic variation with phenotypic variation. They suffer from a number of problems that make it hard to identify specific genes that can be mechanistically (causally) linked to complex traits.

Complex traits are a catch-all term for traits that don't segregate in a Mendelian manner (think punnet squares). Height, which you can measure with absolute certainty, is a complex trait. GWA studies of human height implicate over 700 loci (locations) across the genome.

These loci are correlated with (not causal to) height. This nuance has to do with linkage disequilibrium and the idea that you're mostly finding things near variants that actually might be causal. More importantly, they explain <20% of the known heritability of height.

You can look up 'missing heritability' to read more about this. So if a clearly defined phenotype like height can't yet be broken down into a clear model implicating the coordination of specific causal genes, you can imagine how much more difficult it is with 'intelligence'.

There are also fundamental statistical issues with most human GWA studies, due partly to our inability to do controlled experiments with humans. The statistical thresholds typically reported are artificially relaxed (e.g., not properly controlled for multiple hypothesis testing).

Some, but not all, of these limitations will improve with higher 'n' (more human subjects). But in the end, it is false to report that we have anything but a fuzzy understanding of the basis for intelligence (setting aside philosophical issues as to what 'intelligence' means).