Field: Technology
Genomic Ghosts: Tracing the Mysterious Deep Ancestry Interwoven in Modern Humans
Published | Technical Staff
Visualization
A collaborative team led by researchers at the University of California, Berkeley, has uncovered genetic trails in modern humans pointing to a far more entangled evolutionary history than previously recognized. Using an innovative computational approach, the scientists have found clear evidence for introgression from not only the well-known Neanderthals and Denisovans, but at least two additional, previously unknown archaic hominin lineages—one in Africa, the other in Eurasia—that have left a lasting legacy in the DNA of people alive today.
Until recently, the understanding of human evolution was framed by the relatively sparse fossil and genomic record: high-coverage genomes from only four Neanderthals and two Denisovan individuals, all of Eurasian origin, have dominated models of interbreeding. Most non-Africans today carry 1–2% Neanderthal ancestry; Asian and Oceanian populations bear 0.1–5% Denisovan DNA, reflecting the complex web of admixture across the late Pleistocene. These ancient contributions have been key for adaptive traits—including skin pigmentation, altitude tolerance, and immune defense—but such records overlook epochs and populations outside Eurasia and before 400,000 years ago, due to the absence of ancient DNA.
The breakthrough emerged through a new computational methodology, TRACE (TRacking Archaic Contributions via ARG Estimation). Rather than depending on direct genomic data from fossils, TRACE reconstructs ancestral recombination graphs (ARGs) from hundreds of contemporary genomes, effectively leveraging patterns of linkage disequilibrium and coalescent theory. This method identifies genetic segments whose coalescence times—i.e., the time to most recent common ancestry—are significantly deeper than can be explained by known population splits, flagging putative introgression events from unsampled “ghost” lineages.
Application of TRACE revealed the fingerprint of a previously unknown African “ghost” population. This archaic lineage diverged from Homo sapiens roughly 800,000 years ago, contemporaneous with the split between Neanderthal and Denisovan ancestors. Genomic segments attributed to this lineage comprise about 1% of the genomes of all extant humans, regardless of continental origin—a finding that upends previous hypotheses confining this ancestry to African populations alone. The estimated admixture event predates the primary out-of-Africa migration, occurring over 50,000 years ago. Each individual today harbors approximately 0.5–1% of their genome from this ancient source; these segments are not scattered randomly but often aggregate in functional regions tied to metabolism and immunological response, underscoring the likely adaptive value of archaic alleles.
If the African ghost lineage pushes the horizons of H. sapiens’ reticulate history back by hundreds of millennia, the identification of a second “super-archaic” Eurasian lineage extends the genetic palimpsest even further. TRACE infers that this archaic population split from the main Homo clade more than 1.7–1.8 million years ago—an epoch contemporaneous with, or preceding, the earliest widespread dispersals of Homo erectus. While no direct admixture with modern humans is detected, the super-archaic DNA entered our lineage indirectly: it introgressed into Denisovans during deep Pleistocene times in Eurasia, and subsequently reached H. sapiens via secondary Denisovan admixture events. Notably, the proportion of super-archaic DNA tracks with Denisovan ancestry across modern populations, peaking in Oceanians who harbor the highest fraction of Denisovan heritage.
The combined genetic legacy from these two newly identified lineages constitutes approximately 2% of the modern human genome, a figure that rivals or exceeds the Neanderthal contribution for many populations. The mapping of introgressed segments using TRACE provides a genomic atlas of archaic ancestry. Functional enrichment analyses reveal that many of these archaic-derived haplotypes reside in regions under recurrent selection, including the major histocompatibility complex (MHC), loci relevant to antiviral defense, and genes influencing dietary adaptation. In evolutionary terms, this supports a model where admixture events provided a vital reservoir of standing adaptive variation, subject to subsequent selection in response to novel pathogens and environmental challenges—an embodiment of the “adaptive introgression” hypothesis.
The significance of these findings goes beyond simply cataloging more ancient mixings. In the absence of ancient DNA from most of the world’s continents and epochs, approaches like TRACE enable the reconstruction of deep genealogical history from the mere echoes of structure within the genomes of living people. The method operates by evaluating local genealogical topologies within sequence blocks and comparing their divergence times with expectations under various demographic models. For the African ghost lineage, the excess of long, deeply coalescing haplotypes in both Africans and non-Africans is incompatible with any simple model of Homo sapiens expansion without external admixture; only the insertion of a diverged archaic branch at ~800,000 years ago accounts for both the divergence and the present sharing across continents.
The Eurasian super-archaic admixture, inferred from highly diverged sequence blocks in Denisovan genomes and their derivatives in modern Oceanians, hints at hominins whose existence is not yet attested by fossils or archeogenomes. The possibility exists that this “super-archaic” lineage could be tied to Asian or European Homo erectus populations, whose protein sequences may one day bridge the gap to these missing ancestors. In the meantime, the lingering traces of their DNA in our immune genes and metabolic loci are the only testament to their story.
This research—published July 30, 2026, in Science (Zhang et al., doi:10.1126/science.aef8874)—marks a turning point for paleoanthropology and human genetics. As the world’s genomic reference panels expand in diversity and size, the capacity to unearth additional, even more enigmatic lineages grows stronger. For now, these genomic ghosts—carriers of ancient adaptations and evolutionary gambits—speak silently from within us all, reminding researchers that the saga of human origins is far from complete, and that our very DNA is a living fossil archive, awaiting further excavation through algorithmic ingenuity.
Until recently, the understanding of human evolution was framed by the relatively sparse fossil and genomic record: high-coverage genomes from only four Neanderthals and two Denisovan individuals, all of Eurasian origin, have dominated models of interbreeding. Most non-Africans today carry 1–2% Neanderthal ancestry; Asian and Oceanian populations bear 0.1–5% Denisovan DNA, reflecting the complex web of admixture across the late Pleistocene. These ancient contributions have been key for adaptive traits—including skin pigmentation, altitude tolerance, and immune defense—but such records overlook epochs and populations outside Eurasia and before 400,000 years ago, due to the absence of ancient DNA.
The breakthrough emerged through a new computational methodology, TRACE (TRacking Archaic Contributions via ARG Estimation). Rather than depending on direct genomic data from fossils, TRACE reconstructs ancestral recombination graphs (ARGs) from hundreds of contemporary genomes, effectively leveraging patterns of linkage disequilibrium and coalescent theory. This method identifies genetic segments whose coalescence times—i.e., the time to most recent common ancestry—are significantly deeper than can be explained by known population splits, flagging putative introgression events from unsampled “ghost” lineages.
Application of TRACE revealed the fingerprint of a previously unknown African “ghost” population. This archaic lineage diverged from Homo sapiens roughly 800,000 years ago, contemporaneous with the split between Neanderthal and Denisovan ancestors. Genomic segments attributed to this lineage comprise about 1% of the genomes of all extant humans, regardless of continental origin—a finding that upends previous hypotheses confining this ancestry to African populations alone. The estimated admixture event predates the primary out-of-Africa migration, occurring over 50,000 years ago. Each individual today harbors approximately 0.5–1% of their genome from this ancient source; these segments are not scattered randomly but often aggregate in functional regions tied to metabolism and immunological response, underscoring the likely adaptive value of archaic alleles.
If the African ghost lineage pushes the horizons of H. sapiens’ reticulate history back by hundreds of millennia, the identification of a second “super-archaic” Eurasian lineage extends the genetic palimpsest even further. TRACE infers that this archaic population split from the main Homo clade more than 1.7–1.8 million years ago—an epoch contemporaneous with, or preceding, the earliest widespread dispersals of Homo erectus. While no direct admixture with modern humans is detected, the super-archaic DNA entered our lineage indirectly: it introgressed into Denisovans during deep Pleistocene times in Eurasia, and subsequently reached H. sapiens via secondary Denisovan admixture events. Notably, the proportion of super-archaic DNA tracks with Denisovan ancestry across modern populations, peaking in Oceanians who harbor the highest fraction of Denisovan heritage.
The combined genetic legacy from these two newly identified lineages constitutes approximately 2% of the modern human genome, a figure that rivals or exceeds the Neanderthal contribution for many populations. The mapping of introgressed segments using TRACE provides a genomic atlas of archaic ancestry. Functional enrichment analyses reveal that many of these archaic-derived haplotypes reside in regions under recurrent selection, including the major histocompatibility complex (MHC), loci relevant to antiviral defense, and genes influencing dietary adaptation. In evolutionary terms, this supports a model where admixture events provided a vital reservoir of standing adaptive variation, subject to subsequent selection in response to novel pathogens and environmental challenges—an embodiment of the “adaptive introgression” hypothesis.
The significance of these findings goes beyond simply cataloging more ancient mixings. In the absence of ancient DNA from most of the world’s continents and epochs, approaches like TRACE enable the reconstruction of deep genealogical history from the mere echoes of structure within the genomes of living people. The method operates by evaluating local genealogical topologies within sequence blocks and comparing their divergence times with expectations under various demographic models. For the African ghost lineage, the excess of long, deeply coalescing haplotypes in both Africans and non-Africans is incompatible with any simple model of Homo sapiens expansion without external admixture; only the insertion of a diverged archaic branch at ~800,000 years ago accounts for both the divergence and the present sharing across continents.
The Eurasian super-archaic admixture, inferred from highly diverged sequence blocks in Denisovan genomes and their derivatives in modern Oceanians, hints at hominins whose existence is not yet attested by fossils or archeogenomes. The possibility exists that this “super-archaic” lineage could be tied to Asian or European Homo erectus populations, whose protein sequences may one day bridge the gap to these missing ancestors. In the meantime, the lingering traces of their DNA in our immune genes and metabolic loci are the only testament to their story.
This research—published July 30, 2026, in Science (Zhang et al., doi:10.1126/science.aef8874)—marks a turning point for paleoanthropology and human genetics. As the world’s genomic reference panels expand in diversity and size, the capacity to unearth additional, even more enigmatic lineages grows stronger. For now, these genomic ghosts—carriers of ancient adaptations and evolutionary gambits—speak silently from within us all, reminding researchers that the saga of human origins is far from complete, and that our very DNA is a living fossil archive, awaiting further excavation through algorithmic ingenuity.