SV: [Y-Indology] Indian Express Article today

[Guest guest]

I am not a geneticist, and so, if someone could be kind enough

to explain the following, I'd appreciate it.

 

As usual, the newspapers get it wrong.

A more correct statement of the paper's results

would be that West Asians/Europeans typically

entered Indian society at the higher castes.

 

It is also important to point out that the paper shows

a relationship between two populations; the genetics is

also consistent with a Kshatriya migration out of India.

The direction of movement is inferred from linguistics

and is not independently derived from the genetic data.

 

---

 

Here are a few points that I (a non-geneticist)

found confusing. Explanations from any kind soul

would be appreciated.

 

 

The population sample (for mtDNA tests) was drawn from

Vishakapatnam Dst. The sample was 80 upper caste

(Niyogi and Vydiki Brahmin, Kshatriya, Vysya),

111 middle caste ( Telega, Turpu Kapu, Yadava,

classified as Sudra) and 74 lower (Relli, Madiga,

Mala, classified as Panchama). These are compared to

"Asian", "European" and "African" samples of roughly the

same size.

 

(1) The paper first reports on a "MtDNA HVR1 genetic distance"

which I take to be a gross distance based on mitochondrial

DNA, which is inherited from the mother.

 

In this all castes are closer to Asians than to Europeans

or Africans. Typically, the castes are five to seven times

further away from Africans than from Asians; and are

three to five times further away from Europeans than from Asians.

 

The middle castes are closest to Europeans by this measure,

and are also closest to Asians. If Kshatriya and Vysya are

removed, then the upper caste moves away from Europeans by

a little.

 

(2)The paper then looks at specific sites on the mtDNA ("haplotypes")

 

Here I get a little confused. E.g., the paper says about

haplogroup (set of haplotypes) M is a group found "in

populations that migrated from mainland Asia to Southeast Asia

and Australia, and is found at a much lower frequency in European

and African populations."

 

"Furthermore, these Indian haplogroup-M haplotypes are distinct

from those found in other Asian populations and indicate the

existence of Indian-specific subsets of haplogroup M (e.g., M3)".

 

"As expected if the lower castes are more similar to Asians

than to Europeans, and the upper castes are more similar to

Europeans than to Asians, the frequencies of M and M3 haplotypes

are inversely proportional to caste rank (Table 2.)"

 

(But to me, table 2 does not show this, M and M3 show opposite

trends.)

 

(from table 2)

Incidence of haplogroup

Upper caste M-61% , M3-18.6%

Middle caste M-64.6%, M3-3.5%

Lower Caste M-71.4%, M3-1.4%

 

(3) Most of the Indian West-Eurasian haplotypes "belong to

an Indian-specific subset of haplogroup U, that is U2i,

the oldest and second most common mtDNA haplogroup found

in Europe.

U2i incidence (table 2)

Upper caste 16.9%

Middle caste 9.7%

Lower caste 5.7%

 

But U2i is said to have a coalescence date of 50,000 years

before present.

 

(4) The H,I,J,K,T haplogroups with a more recent

coalescence estimate (how recent is not given) are

five-fold higher in the upper castes than in the lower castes,

6.8% to 1.4% respectively, (but, IMO, are rare).

In the sample, (table 2) H & K was found only in the upper

caste sample, J only in the middle caste sample,

T in middle and lower caste sample only; I not at all.

And, for such small numbers, the standard errors are

large.

E.g., upper caste H is 3.4 +/- 2.0 %,

middle caste J is 0.9 +/- 0.75 %.

 

The significance of this, therefore, escapes me.

 

(5) The paper next discusses at Y-chromosome variations,

these are paternally derived.

 

Here is the Y Chromosome (STRs) Genetic distances

(incidentally, these were done using a different sample

from what is described above).

 

Distance

Caste group Africans Asians Europeans

Upper 0.0166 0.0104 0.0092

Middle 0.0156 0.0110 0.0108

Lower 0.0131 0.0088 0.0108

All castes 0.0151 0.0101 0.0102

 

Notice that if we round up to two places after the decimal,

all groups are equidistant from Asians and Europeans (0.01).

So, the distinctions we are talking about are a few parts

per hundred.

 

(6) The gross comparisons of the Y-chromosome do not yield much,

(as you can see), so the paper goes on to looking at specific

features on the Y-chromosomes.

 

Here is part of table 4 (bi-allelic polymorphisms)

Distance

Caste group Asians Europeans East Europeans

Upper 0.388 0.135 0.073

Middle 0.291 0.146 0.133

Lower 0.376 0.173 0.155

 

This seems unambiguous.

 

(7) Autosomal Genetic distances are given in table 5.

and are consistent with table 4.

 

But, while the Brahmin-European distance is said to be

less than the Kshatriya-European or Vaishya-European distance,

removing Kshatriya and Vaishya from the upper caste group

increases the distance of that group from the Europeans

(from 0.032 to 0.038).

 

So, the burning question for me is -- how were the features

to compare on the Y-chromosome chosen ?

 

(8) Page 6 of the paper : "Analysis of each caste separately

reveals that the genetic distance between the Brahmins

and Europeans (0.013) is less than the distance between

Europeans and Kshatryia (0.030) or Vysya (0.020)."

 

Page 8 of the paper : "This is underscored by the observation

that Kshatriyas (an upper caste) whose members served as

warriors, are closer to Europeans than any other caste (data

not shown)."

 

This suggests to me that a search was undertaken to find the

measures which would show the expected results; not all measures

explored in the research are described in the published paper.

 

-Arun Gupta

[Guest guest]

The paper by Michael Bamshad et al. (available online, for

personal and institutional rs, at the website of

Genome Research, http://www.genome.org/papbyrecent.shtml)

clearly ranks as one of those that has sparked an immediate

interest. It appeared online only on May 8, 2001.

 

There are several interesting features in these findings,

such as the fact that middle castes are closer to Asians

than both the upper and the lower castes, while all three

are equidistant from Africans. On an average, all castes

are slightly farther away from Europeans (genetic distance

of 0.045) than from other Asians (genetic distance, 0.039).

The most interesting finding, however, is the following

(taken from Table 5 of Bamshad et al., p.7).

 

___________

| | Genetic Distance from Europeans |

| Caste |-------------------|

| Group | including Kshatriya | including Kshatriya |

| | & Vaisya in upper caste | & Vaisya in middle caste |

|--------|-------------------------|--------------------------|

| Upper | 0.032 | 0.038 |

| Middle | 0.057 | 0.050 |

| Lower | 0.045 | 0.045 |

|________|_________________________|__________________________|

 

 

For the time being, ignore what these numbers really are, and

also what they really mean. Just take away the idea that the

lower the genetic distance between two populations, the closer

they are related genetically. Also, for the correlations being

done in such statistical studies, models of genetic variability

suggest that at least the third decimal place *is* significant.

One should not be looking at only two decimal places.

 

Now, putting the Kshatriyas and Vaisyas among middle castes,

instead of upper castes, results in:

 

(a) an increase in the genetic distance for upper caste (i.e.

Brahmin only), from 0.032 to 0.038, and

 

(b) a decrease in the genetic distance for middle caste, from

0.057 to 0.050.

 

The reason for this is that Kshatriyas are closer to Eurasian

populations than Brahmins. This is confirmed by an explicit

statement by the authors of the study, in p. 8, that Kshatriyas

are closer to Europeans than ANY OTHER CASTE. Therefore, where

one decides to put the Kshatriyas has a significant effect on

the numbers. The authors thank an "anonymous reviewer" who had

suggested to them that they should look for this effect.

 

I would like to draw the attention of list members to an earlier

posting of mine on the old Indology mailing list -

 

listserv.liv.ac.uk/cgi-shl/WA.EXE?A2=ind0012&L=indology&D=1&P=4045

 

Scroll down to points 3 and 4 in that post.

 

As far as the Bamshad paper is concerned, I would suggest the

following. Genetic distance from Europeans decreases as one goes

higher in caste hierarchy, from zUdra to vaizya and kshatriya.

 

However, it increases again slightly, when one goes higher, from

the kshatriya to the brAhmaNa. In other words, the correlation

is not perfectly linear, from lower castes to higher castes.

There is a curvature at the top, which cannot be lightly ignored

or explained away easily.

 

I would also suggest that one should look at the effect of moving

ONLY Kshatriya data from upper caste to middle caste, instead of

both Kshatriya AND Vaishya. In other words, assign the Vaishyas

always to middle caste, and look at the effect on genetic distance

of moving only the Kshatriya from upper to middle caste status.

 

I would definitely like to see what those numbers would be, for

the sake of objective comparison with currently reported data.

 

Also, I would like to see the following numbers, independent of

comparison with contemporary Asian and European populations -

 

genetic distance between brAhmaNa subjects and lower castes,

as measured by both maternal mitochondrial DNA and paternal

Y-chromosome inheritances, and

 

genetic distance between kshatriya subjects and lower castes,

as measured by both maternal mitochondiral DNA and paternal

Y-chromosome inheritances.

 

This would have obvious implications for the actual incidence

of both theoretically sanctioned hypergamy and theoretically

prohibited varNa mobility in India.

 

Finally, FYI - The study was done in collaboration with Andhra

University, and it looked only at a Telugu speaking population.

There is another paper by Bhattacharya et al. published in the

same journal (Genome Research, 1999, vol. 9, pp. 711-719), by

researchers from Calcutta, including people from the Indian

Statistical Institute. This study was based on populations from

Orissa, West Bengal and Uttar Pradesh, and showed that there is

negligible male gene flow across ethnic boundaries (caste and

tribal, but not geographical location) in that part of India.

The most striking result of that study, however, was "... Y-

chromosomal variation is not structured by social rank...".

 

Best regards,

Vidyasankar