Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 173 (1), 53-61.e9

Analysis of Human Sequence Data Reveals Two Pulses of Archaic Denisovan Admixture


Analysis of Human Sequence Data Reveals Two Pulses of Archaic Denisovan Admixture

Sharon R Browning et al. Cell.


Anatomically modern humans interbred with Neanderthals and with a related archaic population known as Denisovans. Genomes of several Neanderthals and one Denisovan have been sequenced, and these reference genomes have been used to detect introgressed genetic material in present-day human genomes. Segments of introgression also can be detected without use of reference genomes, and doing so can be advantageous for finding introgressed segments that are less closely related to the sequenced archaic genomes. We apply a new reference-free method for detecting archaic introgression to 5,639 whole-genome sequences from Eurasia and Oceania. We find Denisovan ancestry in populations from East and South Asia and Papuans. Denisovan ancestry comprises two components with differing similarity to the sequenced Altai Denisovan individual. This indicates that at least two distinct instances of Denisovan admixture into modern humans occurred, involving Denisovan populations that had different levels of relatedness to the sequenced Altai Denisovan. VIDEO ABSTRACT.

Keywords: Denisovan; Neanderthal; ancient admixture; archaic introgression; out-of-Africa migration.


Figure 1
Figure 1. Comparison of our method (Sprime) with a previous method (S*) on simulated data
Detection frequency is the proportion of haplotypes with detected introgression after removing false positive results, and has a maximum possible value of 0.03 (the simulated admixture proportion). Accuracy is the proportion of the putative introgressed alleles that are truly introgressed. We use data simulated with a constant recombination rate (ρ = 10−8) and constant mutation rate (μ = 1.2 × 10−8) so that the application of a score threshold is equivalent to application of a p-value threshold. The full simulation model can be found in Methods. We show results for a range of score thresholds for each method. The default threshold for our method is 150,000 (black data points). Analyses were applied to 100 simulated target individuals and 100 outgroup (African) individuals, except as otherwise noted. Simulated regions were 10 Mb long. Analyses with S* used a sliding window of 50 kb, moving by 10 kb each step. The standard analysis with our method analyses the full region without a sliding window, but we also applied it with the 50 kb sliding window for comparison. The 2014 variant of S* analyzes target individuals in subsets of 20, and uses only 13 of the outgroup individuals. For comparison we also show results for Sprime with 15 target individuals. The 2016 variant of S* analyzes target individuals one by one (utilizing only the two haplotypes within an individual to determine LD), and uses all 100 outgroup individuals. See also Figures S1 and S2.
Figure 2
Figure 2. Tiling of introgressed haplotypes
This example is from analysis of Utah residents (CEU) from the 1000 Genomes Project. A single inferred segment (i.e. tiling of overlapping putatively introgressed haplotypes) is shown. This segment includes 141 putatively archaic-specific variants, with a match rate of 94% to the Altai Neanderthal genome. Each horizontal line represents a introgressed haplotype, with the line running from the start of inferred introgression to the end of inferred introgression. Many haplotypes share the same start or end point due to inheritance of that part of the segment from a common ancestor.
Figure 3
Figure 3. Detection and match rates in 1000 Genomes populations
(A) Match rates are for the Altai Neanderthal. (B) Match rates are for the Altai Neanderthal or Altai Denisovan. Population codes can be found in Table 1. Populations are colored by region (European in black, East Asian in red, South Asian in blue, American in magenta). Match rate is the rate at which putative archaic-specific alleles match the sequence archaic genomes. Detection rate is the average proportion of each haplotype inferred to be introgressed. See also Figure S3.
Figure 4
Figure 4. Contour density plots of match proportion of introgressed segments to the Altai Neanderthal and Altai Denisovan genomes
The match proportion is the proportion of putative archaic-specific alleles in a segment that match the given archaic genome, excluding alleles at positions masked in the archaic genome sequence. Segments with at least 10 variants not masked in the Neanderthal genome and at least 10 variants not masked in the Denisovan genome are included. Numbers inside the plots indicate the height of the density corresponding to each contour line. Contour lines are shown for multiples of 1 (solid lines). In addition, contour lines for multiples of 0.1 between 0.3 and 0.9 (dashed lines) are shown for additional detail. European populations are given in the first row, East Asian populations in the second row, South Asian populations in the third row, and American and SGDP Papuan populations in the final row. Additional information about the populations can be found in Table 1. See also Figure S4.
Figure 5
Figure 5. Mean amounts of detected introgressed material per individual, classified by affinity to the Altai Neanderthal and Altai Denisovan genomes
Definitions of the affinity groups are given in Methods. Unclassified material includes segments that are too short to be confidently classified into an affinity group, as well as longer segments that have low levels of affinity to the archaic genomes.

Comment in

Similar articles

See all similar articles

Cited by 21 articles

See all "Cited by" articles

Publication types

LinkOut - more resources