NEW YORK (GenomeWeb) – A new analysis suggests there were two instances of Denisovan admixture into modern humans.
Archaic hominins have left their mark on the modern human genome, as present-day Eurasians can trace about 2 percent of their genomes to Neanderthal origins, while Oceanians have inherited about 5 percent of their genomes from Denisovans.
Researchers from the University of Washington and Princeton University applied a new reference-free approach to uncover introgressed haplotypes within modern human genomes. As they reported in Cell today, their analyses found that the Denisovan ancestry in East Asians originated from a different Denisovan population than the Denisovan ancestry found in South Asians and Papuans. It had previously been thought that Denisovan ancestry in Asians had been acquired through the Papuan population.
"In this new work with East Asians, we find a second set of Denisovan ancestry that we do not find in the South Asians and Papuans," UW's Sharon Browning said in a statement. "This Denisovan ancestry in East Asians seems to be something they acquired themselves."
She and her colleagues developed a new approach, called Sprime, to tease out archaic introgressions from large-scale genome-wide data. This method, which is freely available as a software package, is similar to S*, in that it relies on the unusual mutational characteristics of introgressed haplotypes to differentiate them from other sequences. According to the researchers, Sprime had increased detection power as compared to versions of S* and had an accuracy of at least 93 percent across a range of simulated samples with difference divergence and admixture times.
They then applied this approach to analyze the non-African populations from the 1000 Genomes Project dataset, as well as the UK10K dataset and Papuans from the Simons Genome Diversity Project, a total of 5,639 whole genomes. As expected from previous studies, East Asians had higher introgression rates than Europeans, while South Asians and Europeans had similar introgression rates.
The researchers then compared the putative archaic alleles their approach inferred to actual sequenced archaic genomes. For Europeans, for instance, there was about a 0.7 match rate with the sequenced Altai Neanderthal genome.
When they then examined how well those inferred positions matched either Altai Neanderthal or Altai Denisovan sequences, Browning and her colleagues found a large cluster matching Neanderthals and a smaller one matching Denisovans most global populations tested. But in Asian and Papuan populations, they noted a third cluster that was more similar to the Altai Denisovan population than to the Altai Neanderthal population.
Within Japanese and Chinese populations, the researchers noted that the Denisovan cluster of segments had a bimodal distribution of match rates to Altai Denisovan sequences, with one cluster more closely related to Altai Denisovan sequences than the other. As a test for two distinct sources of Denisovan ancestry was statistically significant, this indicated to the researchers that two different Denisovan populations admixed with Asian populations.
Browning said this raised the possibility that a more southerly population of Denisovans admixed with the ancestors of Papuans, and a more northerly group admixed with the ancestors of East Asians.
This finding of two instances of Denisovan admixture indicates that there might be additional interbreeding events to tease out. "We want to look throughout the world to see if we can find evidence of interbreeding with other archaic humans," Browning said. "There are signs that intermixing with archaic humans was occurring in Africa, but given the warmer climate no one has yet found African archaic human fossils with sufficient DNA for sequencing."