From Dhu on Reich's paper, cut-pasting from IF:
Posted 06 October 2009 - 11:17 AM
The entire narrative is crafted to scuttle the southern route. Fact of Southern route coupled with a delayed (by 30-40K) colonization of West by AMH means that India is automatically the source of European lines. The Basque intellectual:
"There's a new seemingly important paper around but it is behind a paywall so I can only make a shallow comment: David Reich, Kumarasamy Thangaraj, Nick Patterson, Alkes L. Price & Lalji Singh, Reconstructing Indian population history. Nature, 2009."
A divulgative review can be read at Science Daily.
Apparently the research has found or confirmed two important elements in the Indian genetic background:
1. That there were two ancestral populations in the subcontinent: a northern one (ANI), not very different from Western Eurasians and a southern one (ASI) totally unique to India and whose only pure remainder are the Andamanese. Otherwise both distinct ancestral populations are nowadays found mixed at various apportions.
2. That tribes and castes are really not that different, with castes seemingly being formed out of recycled ancient tribes. This means that there are many tribal/caste founder effects that make each of these groups rather unique and inbred, with likely relevant health consequences like those found among other similarly isolated populations, such as Jews or Finns.
Nothing too new, because I do have the feeling that this was vaguely known but never really established in such categorical terms. For example, in some previous global studies South Asians have often appeared as a mixture of two components: one the same as West Eurasians and another unique of them. Even the anthropometrists of old used to talk of "Caucasoid" and "Australoid" (or sometimes "Veddoid") Indians, even if we know now that the use of the term "Australoid" in the past was all but clarifying, mostly meaning anything Eurasian that is not specifically Western or Eastern.
For me, this confirmation of the existence of two distinct ancestral populations, suggests that they formed in the early period of Eurasian spread of humankind, before what we conventionally call now "races" formed. As West Eurasia was colonized since c. 50,000 BP, that provides a most recent time limit for such population divergence, because it must have happened before, probably quite a bit of time before, some of the ANI peoples migrated to West Asia, North Africa and Europe. The "Indian remix" must then have happened after this westward migration.
---
Posted 03 January 2012 - 07:01 AM
We know from the haplogroup studies that the entire "West Eurasian" genome is derived from India. At the level of mtDNA, the N and U types found in W. Eur were wintering in the Indus region prior to AMH's first venture to the West. From Metspalu it is clear that these groups have very low autosomal diversity in comparison to populations east. this fact has interesting consequences for the intermediate diversity component (K5) found in the west: it cannot have been generated in regions through which the low diversity WE groups necessarily resided and must signal secondary flow from regions further east. Supportive is Metspalu's evidence regarding the very weak cline of K5 in India from the Indus basin, versus the robust cline westwards including across the refugia of the Caucasus:
"However, considering the geographic spread of this component within India, there is only a very weak correlation (r = 0.4) between probability of membership in this cluster and distance from its closest core area in Baluchistan. Instead, a more steady cline (correlation r = 0.7 with distance from Baluchistan) of decrease of probability for ancestry in the k5 light green ancestral population can be observed as one moves from Baluchistan toward north (north Pakistan and Central Asia) and west (Iran, the Caucasus, and, finally, the Near East and Europe)."
Probably it was this intermediate component which pushed out the initial low diversity components (which themselves could not have arisen anywhere else but at the edges of S Asia) to the West in the first place. Additionally, the cline runs *across* the Caucasus from an immediate origin in C Asia. The C Asian component in turns seems to derived from the Makran coast/Baluchistan/Sindh (immediately West of the Indus) and is the logical and evidenced origin point for these migration types. The Sindh> C Asia> West scenario was even proposed by Elst as generally characteristic of the out-of-India route:
"A look at the map suffices to show the improbability of any other route from India to Iran: rather than to go in a straight line across the mountains, substantial groups of migrants would follow the far more hospitable route through the fertile Oxus valley to the Aral Lake area, and then proceed south from there."
So we have in Metspalu a very basic genetic parsing, reflecting these geographical divisions and as formed by the out-migrations.
Any contrary scenario (out of Caucasus for example) would have the intermediate component K5 going against the "diversity cline" (with S Indian K6 as the most diverse) into the NW region of the subcontinent which we know independently from mtDNA evidence had harbored an "effective population size" of considerable antiquity (see Atkinson, Gray, Drummond
http://mbe.oxfordjou.../468.abstract). It is quite obvious that such a scenario would be near impossible especially without producing any appreciable cline in India. To counter this fact of paleolithic time ranges required for the generation of diversity (as required by the highest diversity component in Paleo S India), the familiar elite and quasi-monotheist scenarios need to be brought in.
More interestingly, the K5 components in Russia and France appear to have directly come from S Asia rather than the Caucasus and it is this specifically which is the out-of-india IE component and corresponds to the now well-attested r1a1 migration westwards.
What we are seeing with the autosomal evidence of an India-specific group is merely the restrictedness of mtDNA macro-haplogroup M to India and further East. Of course this does not mean that N and R did not arise in India, they most definitely did as well. It is a different matter, that N and R can be *labeled* as "West Eurasian" and the aryans can thus be ushered in against all phylogeny- and trajectory-based analyses.
The Reichs and even Metspalus are wedded to the idea of an implied separate origin for WE phenotype being a proxy for AIT. One of Reich's follow-up papers even posits an acceleration of ANI/ASI interaction at 1500 BC and this is supposed to be the AIT!! Romero's LP paper required an Aryan reading of Reich's ANI to "determine" LP introgression into S Asia from the West, overlooking the fact that LP is highly selective as a Vitamin D substitute away from the tropics. The high allele frequency in Europe would be significant only if it were not for an highly selective trait and in such a case the locus would by diametrically opposite to the region of selection! From Reich's and Romero's follow-up papers, it is quite clear that they were using errant and normative reading of 'terms' designated for the paleolithic scenario to *re-interpret* problematic data. I counted 3 non-trace unique LP haplotypes for Pakistan, 1 non-trace and 3 trace for the Near East, and 1 trace for Europe in Romero's Table 3. But in spite of the allele diversity data available, Romero and Reich have used ANI plausibility arguments to "determine" introgression against the selection cline (LP is not selective in Pakistan/NE)
Apart from Reich's normative use of the term ANI, another trick is to use the artifactual clines formed by divergence isolation to overrule the informative diversity clines. Also used is Fst closeness among the migrated out populations and to use intermediate positions on PCA for the source/transit populations (intermediate positions are seen when the migrations are into a well established populations, e.g., in the "admixed" Tajiks compared to the WE-tranformed Turkmen in the case of migrating Turkics; i.e., Turkmen are Tadjiks who have completed the transit).
Even non-wave/non-phylogenetic (non-contextual) Parsi and sea-borne, Roman, British, and Pondicherian era haplogroups are being recruited for the cause. For the latter two, the monotheist elite paradigm is obvious while the first was a fugitive which was sheltered. Significantly, for the Pondicherian, there is a confirmatory mirror in European colonization of the Caribbean. The recent TMRCA for the Pondicherian, and in more general terms for monotheism itself, is one of the limiting criteria for AIT date speculation.
R is anchored in interior Asia if not by anything (due to coeval spreads), then by P definitively. Nonetheless, asiatic R1a1 intrudes into Europe and forms a distinct 2nd wave; European R1a1 which is under one or two SNPs and is delimited to Europe is almost definitely derived from a 1st wave into Europe from the same eastern source: the definite clue is the known trajectory of the 2nd wave intruding into Europe from the East . This dovetails perfectly with Talageri's well-argued theses for the 2nd wave: placing the Iranians out of Kashmir onto the Parusni, clashing with the interiors on the Saraswati, and then out into C Asia and into Iran (also Mittanis, Kassites trajectories; Semitic/Hurrian preponderance in the NE, etc). Via Nichols, we know that the combined Indo-Iranian homeland is also the IE one. The wave trajectory from Asia is now well known for R1b, and the same also holds true for R1a1.
---
A Debate in Sulekha discussions boards on Reich and his Indian associates