The following are the Y-DNA haplogroups of the mainstream eurasian males.
>G,H,I,J,K
among which G,H,I,J are thought to be of Caucasoid. (as a side note, you'd all know that modern caucasoid haplogroup R, the speakers of proto-IE language, was the descendants of K, which is of broadly east eurasians including pacific islanders and mongoloids.)

here, let's have a closer look at the languages attributed to the respective haplogroups.
G : kartvelian (ergative languages)
H : dravidian (some isolated ones show split ergativity)
I : basque (ergative language)
J : elamite (ergative language)

>>17832619 (OP)
E is west eurasian too

>>17832631
I know. but it would be the same to R.

>>17832636
E come from MENA then got into africa and europe
Not comparable to R

>>17832651
okay. it's not the point now anyway.

>>17832619 (OP)
After about 10,000 years of separation under pre-literate conditions, natural drift makes it impossible to trace any genealogy of languages. Since the separation of these y-dna lineages dates back multiple tens of thousands of years, your question is moot.

>>17832681
depends.
at least the kartvelian, dravidian, and basque are thought to have been very indigenous to their lands.
only the elamite language is unclear about its origin among the examples.

bump

>>17832619 (OP)
I don't understand what ergativity means and whether it's a real distinction so it can be used to show common descent or just made up by linguists. But Sumerian and Hurro-Urartian are also apparently ergative so looks like that's the case.
Your haplogroup divisions are wrong though. There were maybe 2 distinct population groups in the middle east (one with G and H2 and another with J and LT). I would have left to Europe between 40 and 30 thousand years ago (although WHG have extra Anatolian from around 20 kya). Indian H1'3 separated from H2 around the time of out of Africa, but Dravidian is probably from recent Iranian migrants anyway (who brought J and L).