I have found this piece very informative as the information it contains is from India's premier institute of Genomics. I have resisted adding any of my comments, it's quite fascinating. Tho Book is India - A Portrait by Patrick French published in 2011:
The campus of the IGIB, India’s secretive Institute of Genomics & Integrative Biology, is set in a leafy part of north Delhi.
Dr. Arijit Mukhopadhyay, “My interest is in RGC death,” he said. “RGC [retinal ganglion cell] death is the final cause of glaucoma, a disease of the visual nerves culminating in blindness.”
Mukhopadhyay was one of only two people authorized to speak about an extraordinary project that was under way here. A consortium of mainly young scientists was attempting to unravel the complete genetic map of the Indian peoples. It was an ambitious concept in a country with over one billion inhabitants, several thousand endogamous groups (who married only within their community) and more than 300 distinct functioning languages. The project had been conceived by Professor Samir Brahmachari, a bio-physicist, and from the start he had realized he was stepping on dangerous ground. The information they were liable to discover about the origins of communities might have political, religious or caste consequences, and if mishandled could lead to conflict and even violence. The media had already run several stories which, in the view of the IGIB, distorted their research.
“We are trying to draw a genetic landscape of India and use it as a canvas to identify disease genes and genetic markers,” said Dr. Mukhopadhyay cautiously from behind his desk. “We will learn which groups are prone to particular diseases or do not react to certain drugs. We are in the early days of an idea that has huge implications. We have looked at fifty-five populations so far, across the length and breadth of the country. This is the discovery stage. For example, we found that susceptibility to HIV is lower in a certain group in the north-west of India. A particular protective form of the gene has stayed within that population, because they are endogamous. The more outbred their community becomes, the more it will spread.”
He showed me a bar chart and a map of India covered in circles of different colours, some like pie charts.
“You can see here the Indian peoples fall into four major language families: Indo-European, Austro-Asiatic, Tibeto-Burman and Dravidian.” The first group spread across the upper and middle reaches of the map, the Austro-Asiatics (the tribals or Adivasis) were clustered in the centre and east, the Dravidians covered the south as far up as Andhra Pradesh and the Tibeto-Burmans were confined to small border areas to the north.
“Now look at these. This stuff was very unexpected.” He pointed at the coloured circles on the map. “You will see that Indians are more similar than you would think across the country. There are membership exchanges within these four groups. It’s all mixing up, even with the Dravidians.”
Now I understood the circles: they showed genetic groups where you would least expect to find them. In Kashmir, there were people who were genetically similar to Dravidians. In Gujarat and eastern Orissa the same was true, though the markers were weaker. In parts of Andhra Pradesh and Karnataka in the south of the country, the situation was reversed: here were Indo-Europeans. In Madhya Pradesh in central India, you could find groups whose ancestors had come from the Himalayas. So people who thought they were a product of a place where their family had lived for infinite generations were genetically closer to Indians who lived a thousand miles away, and spoke a different language.
When I pointed to a spot on the map and asked Dr. Mukhopadhyay exactly where it was, he hesitated.
“We don’t say the place names. We agreed, because of the political risks, not to release the ID of the blood-sampled groups—Hindus, Muslims, Sikhs. This is uncomfortable territory.” He touched two red circles near Pakistan. “They probably would not eat in one another’s home. How deep are our genetic relationships, and yet how different are our social relationships. The cultural structures we are following are new: it takes time for practices like not marrying into another community to come out genetically. It takes a few thousand years.”
Tens of thousands of years?
“No, thousands of years.”
So with all this mixing and complexity, could you test for caste?
“There is no scientific basis to say you could have a caste gene. For a start, in our research we use samples of fifty or a hundred people, not individuals. If you test a population group in India and look at twelve genetic markers—DNA sequence variations—you have nearly a 100 percent chance of knowing if they are tribals or not, and an 85 percent chance of establishing their language group. The data would not tell you the caste, because there is no basis. You can’t say who is ‘superior’ or ‘inferior.’ I travelled to Pune to take some blood samples and one person—I should tell you that he was a political type—said to me: ‘You can take my sample, and prove that I am the supreme quality of human in India.’ There’s a strange dynamic in our society.”
Dr. Mukhopadhyay had to leave for a meeting and introduced me to his colleague Dr. Mitali Mukerji. She was the only other person authorized to speak about the project.
“It seems like a lot of Bengalis work here,” I said.
Dr. Mukhopadhyay smiled. “I am a native of Calcutta. If a job is advertised, seven out of ten applicants are Bengali. Some say, ‘Ah, Bengalis are more clever because they eat a lot of fish and get omega-3 acids.’ I tell them: it’s not like that, clever Bengalis go to academia and clever north Indians go to commerce.”
Dr. Mukerji specialized in molecular genetics and coordinated the project across six institutes in the country. “I considered working abroad, but there was a lot to do on genomics here,” she told me. “We were studying the history of diseases and realized we needed a proper genetic landscape map of India to do our work. Take Huntington’s disease as an example. It is caused by DNA mutation and can get worse with each generation. It’s very rare in Asia. It came to India by two routes, one from the north and the other probably through Mysore at around the time of Tipu Sultan. There is evidence that the form of Huntington’s disease which we see in south India originated in Ireland, and that it probably came from soldiers who were made captives by Tipu Sultan.”
I was still digesting what I had heard about genetics in India and asked Dr. Mukerji what had surprised her most during the ongoing research project.
“We soon realized there was not a prototype Indian. People appear to have come in from places that are now Iran, Burma, central Asia, Afghanistan. This supports the idea of waves of settlers from various directions. There are many admixtures, whereas by comparison Caucasians are homogenous. India is like a melting pot compared to other Asian countries. If you trace mitochondrial DNA, it shows that women moved around and probably reproduced with other communities. Marriage within your group is more recent in India.”
What she said was not good news for racial theorists or for Indians who saw their position in the hierarchy of caste as being ordained by scripture and tradition. Had Kshatriyas not married Kshatriyas and Brahmins married Brahmins forever? Hinduism does not depend on a linear concept of time and has no single sacred book—it is not a “revealed” religion like Christianity, Islam or Judaism—but it has an overwhelming sense of its own antiquity. If you leave aside for a moment the influential reform movements of the nineteenth and twentieth centuries, which sought to strip out distractions such as caste from religion, most variants of Hindu practice stress that rituals have been done by Brahmin priests in a particular way for aeons. If the ancient social order was not divinely ordained and has not existed since the dawn of civilization, it was invented by humans for a social purpose. This has been a progressive point of view for many years, but the IGIB project threatened to give it a proof and validity.
In the months after my visit to the IGIB, rival scientific reports were published in different parts of the world, including a study which claimed, on the basis of a much smaller sample survey, that India’s caste groups had been endogamous for millennia. This was widely reported, and any future research which appears to show that caste is genetically identifiable will certainly receive publicity. The notion of such a root distinction appeals both to upper-caste traditionalists and to lower-caste political leaders who claim they represent the descendants of India’s original inhabitants (whom they imagine were beaten down and exploited by fair-skinned Aryan invaders).
Dr. Mukerji stressed the project was by no means complete. “There’s so much we are still discovering. Skin pigmentation is only the property of a few genes, so you just need to flip some genes to get from dark to light skin. It will be some years before we understand this fully. I was surprised that people we tested from Ladakh, Sikkim and the north-east were so similar, and that some Kashmiris may originally have been Dravidians who were pushed up to the north. Upper-caste Hindus seem to be much closer to Muslims than to the high castes from other places. We need to study in more detail in the south, where the communities have remained more distinct. Adivasis are noticeably different. They may be connected to Australian aborigines.”
“Does the genetic landscape map reveal anything about caste?” I asked. “Some people tell doctors that they like sperm donors to come from their own caste, to get the right genes.”
Dr. Mukerji was dismissive. “There’s no logic to talking about caste and sperm and which community has better genes. Indians all have opinions, but the caste system has no genetic basis.”