Dienekes got his hands on Otzi's genome finally, and decided to confirm some suspicions. In general no great surprise, though I think the number of SNPs he used (44,000) is a little on the low side for the questions he was asking. But the details here aren't too relevant because all the available evidence points to the "Iceman" being affiliated with modern day Sardinians, of whom we know much more with many more markers. In any case, he points out that if you run ADMIXTURE you tend to see that Sardinians, and to a lesser extent Basques, are lacking in some ancestral components. One phenomenon which is implied by this is that these populations which are less "cosmopolitan" may reflect more ancient patterns, when there was less admixture. I've indicated this myself when it comes to non-Muslim minorities in the Middle East. But one caution I would immediately make is that we are judging the variation of smaller populations by the yardstick of larger populations. Distinct and less numerous groups, such as Sardinians, may show less ancestral cosmopolitanism in part because the reference populations which could be used to adduce such a state no longer exist. To give an extreme example the Onge of the Andaman Islanders often pop up as a very distinctive genetic component. But what if there were many more related populations in the data sets generating patterns of variation? We might see that the Onge themselves are composites! The idea I'm trying to get across is that we imagine the past was demographically pristine. But if it wasn't, then our attempts to make inferences become all the more difficult.