If the common mutation of SLC24A5 confers an advantage other than increased vitamin-D production, evidence might show up in a mixed ethnic group, formed from one population with a high incidence of the interesting SLC24A5 variant and another with little or none. You could look at the overall admixture fraction – say, 50% African and 50% non-African – and see if the SLC24A5 variant is more common today than you’d expect from the original admixture proportions (which will be easily determined from looking at neutral loci).
This has already been done for Ethiopia. There, in the Semitic and Cushitic populations, where the estimated proportions of African and non-African ancestries are approximately equal, SLC24A5 is significantly more common than you would expect.
Those populations live up on the plateau, over a mile high and within ten degrees of the equator. There’s enough UV to roast a goat. Somehow I doubt if Vitamin D is paying the freight here.
As for those who assume that sexual selection must be driving that increase – show me the time machine. I don’t know if there was any such preference over the past three thousand years in Ethiopia and neither do you. This is used as an excuse to avoid looking at the biochemical details and trying to find out what’s actually happening. If I hear it again, I may have to call the elephants.
Next, someone should do the same for the Deccan plateau, which ought to be easy.