As it happens, there's recently been a notable study of native american genetics.
The expected bits:
- the populations of america and the pacific northeast (chukchi, etc) all result from various mixtures of siberians with east asians
- almost all americans descend from a single (60% east asian, 40% siberian) population in Beringia
- the na-dene samples were american with a dash of asian (90-10). This doesn't seem surprising - while the na-dene have long been considered linguistically and culturally distinct, a later migration, most athabaskans have been known for a while to be pretty amerindian in genetics. This suggests a culturally-influential migration that gradually intermingled with the much-more-numerous natives
- greenlanders are 2/3rds Saqqaq (one of the pre-Inuit 'Eskimo' groups), and 1/3rd american. Again, it makes sense that the initial pre-eskimo invasion would mingle with the natives, and that the result would be the source of the second wave
- there was at one stage a relict population in Alaska, from which all the other americans split off (i.e. the people who stayed when the pioneers went south), but which has at some point been lost. Given the harsh conditions, multiple invasions from asia and possibly back-migration from further south, this isn't a surprise.
- the study splits americans into 'north' and 'south' as the primary split in the modern population. However, their "south american" consists of one Aymara sample and one from Montana. Nonetheless, other studies have repeatedly confirmed the basic north/south split - though more studies would be required to see how universal it was. The fact that the Clovis Culture sample from Montana patterns with south america seems to suggest it's not just a simple "inland north america vs coastal south america" split we're looking at, though.
- the Asian componant of na-dene isn't particularly closely related to the asian componant in Ket. The closest connection was to Koryak, then to Saqqaq, THEN to Ket. [the siberian admixture into Ket seems even more divergent from that into everything pacificky, so was probably later]. This may mean:
-- Dene-Yeneseian is wrong
-- The Ket have been genetically replaced by another siberian/asian mix in the same area but wih different ancestry, while keeping their language
-- The Na-Dene have been genetically replaced twice (first by a different asian group, and then by americans)
-- Not only are EA and CK related to DY, but they're actually internal to DY somehow
-- There was once a more widespread family comprising Dene-Yeneseian and the former languages of the Koryak and Saqqaq, but the latter groups adopted the languages of CK and EA groups later on.
Impossible to say at this stage which it is, I think.
But the river tripped on her by and by, lapping
as though her heart was brook: Why, why, why! Weh, O weh
I'se so silly to be flowing but I no canna stay!