The mixing of two distinct lineages led to most modern-day Indians.
Elie Dolgin
Most Indians are a mix of two ancient lineages.Getty
The population of India was founded on two ancient groups that are as genetically distinct from each other as they are from other Asians, according to the largest DNA survey of Indian heritage to date. Nowadays, however, most Indians are a genetic hotchpotch of both ancestries, despite the populous nation's highly stratified social structure.
"All Indians are pretty similar," says Chris Tyler-Smith, a genome researcher at the Wellcome Trust Sanger Institute near Cambridge, UK, who was not involved in the study. "The population subdivision has not had a dominating effect."
India makes up around one-sixth of the world's population, yet the South Asian country has been sorely under-represented in genome-wide studies of human genetic variation. The International HapMap Project, for example, includes populations with African, East Asian and European ancestry — but no Indians. The closest the Human Genome Diversity Cell Line Panel of 51 global populations comes is Pakistan, India's western neighbour. The Indian Genome Variation database was launched in 2003 to fill the gap, but so far the project has studied only 420 DNA-letter differences, called single nucleotide polymorphisms (SNPs), in 75 genes1.


Caste divisions
Now, a team led by David Reich of the Broad Institute in Cambridge, Massachusetts, and Lalji Singh of the Centre for Cellular and Molecular Biology in Hyderabad, India, has probed more than 560,000 SNPs across the genomes of 132 Indian individuals from 25 diverse ethnic and tribal groups dotted all over India2.
“There are populations that have lived in the same town and same village for thousands of years without exchanging genes.”
David Reich


Broad Institute
The researchers showed that most Indian populations are genetic admixtures of two ancient, genetically divergent groups, which each contributed around 40-60% of the DNA to most present-day populations. One ancestral lineage — which is genetically similar to Middle Eastern, Central Asian and European populations — was higher in upper-caste individuals and speakers of Indo-European languages such as Hindi, the researchers found. The other lineage was not close to any group outside the subcontinent, and was most common in people indigenous to the Andaman Islands, a remote archipelago in the Bay of Bengal.
The researchers also found that Indian populations were much more highly subdivided than European populations. But whereas European ancestry is mostly carved up by geography, Indian segregation was driven largely by caste. "There are populations that have lived in the same town and same village for thousands of years without exchanging genes," says Reich.


Number puzzle
Indian populations, although currently huge in number, were also founded by relatively small bands of individuals, the study suggests. Overall, the picture that emerges is of ancient genetic mixture, says Reich, followed by fragmentation into small, isolated ethnic groups, which were then kept distinct for thousands of years because of limited intermarriage — a practice also known as endogamy.
This genetic evidence refutes the claim that the Indian caste structure was a modern invention of British colonialism, the authors say. "This idea that caste is thousands of years old is a big deal," says Nicole Boivin, an archaeologist who studies South Asian prehistory at the University of Oxford, UK. "To say that endogamy goes back so far, and that genetics shows it, is going to be controversial to many anthropologists." Boivin fears that the study might be 'spun' by politicians seeking to maintain caste structures in India, and she calls on social scientists and geneticists to collaborate on such "highly politicized" issues.
Beyond the study's social repercussions, the low rates of genetic mingling "could have important implications for biomedical studies of Indian populations", notes Sarah Tishkoff, a human geneticist at the University of Pennsylvania in Philadelphia who was not involved in the research. The partitioned population structure will need to be taken into account in any efforts to map disease genes, she says.
The small numbers of founders of each Indian group also have clinical consequences, says Reich. "There will be a lot of recessive diseases in India that will be different in each population and that can be searched for and mapped genetically," he says. "That will be important for health in India."
The evidence that most Indians are genetically alike, even though anthropological data show that Indian groups tend to marry within their own group, is "very puzzling", says Aravinda Chakravarti, a human molecular geneticist at the Johns Hopkins University School of Medicine in Baltimore, Maryland, who wrote an accompanying News & Views article3. For example, Chakravarti notes that the study can't establish a rough date for when the ancient mixing between the two ancestral populations took place. "There are very curious features of the data that are hard to explain," he says, adding: "This is not the end of the story."


References
Indian Genome Variation Consortium J. Genet. 87, 3-20 (2008).
1. Reich, D. et al. Nature 461, 489-494 (2009).
2. Chakravarti, A. Nature 461, 487-488 (2009).