Recently more N2a full genome results became available, as well as a few W outliers, some of them from 23andme rather than full genome tests. When processing these with the Fluxus network software, an interesting family tree appears. In this version of the W family tree, W1 is ancestral to W3,4,5,6, rather than being a subgroup of the common 'W' type. Phylotree.org can't accept this conclusion yet, since by their correctly rigorous standards multiple credible full-genome results have to be published on genbank for each branch of the tree before it can be accepted as valid. But it is an interesting possibility, that will require the rewriting of group W ancestry, if further results substantiate it:
Comments? Corrections? Questions? E-mail me!