With all the talk about Basques I decided to do my own analysis with Admixture. Dienekes gave me a copy of his IBS file, which has all the 1000 Genomes Spanish samples, including Basques. I merged it with the HGDP sample, which has French Basques (just “Basques” in the plots below) and French non-Basques. I pruned most of the populations, but kept the Mozabites, which are a Berber group from Algeria. The number of markers was ~350,000, and I ran it up to K = 8, or 8 component populations. I stopped there because the components were starting to break up in a very choppy manner.
In general I do think that the idea that non-Basque Spaniards have Moorish genetic input seems supported. It isn’t definitive though. And you have to be careful, there are lower parameter values where Sardinians seem to have an affinity with Mozabites to a great extent, even more than Spaniards. But that disappears as you move up the number of K’s. But who is to say which K is the correct K? The consistent Sub-Saharan African among non-Basque Spaniards (also evident in the Behar et al. data set) component probably convinces me that there was a Moorish impact, since these are likely to have come with the Islamic conquest, and not Phoenicians.
All the files from the Admixture run (and csv files with tabular results) are here.