Ghost Lineage in Africa Contributed 1% of Human DNA

A third archaic lineage hides in our genomes

New research published in Science (DOI: 10.1126/science.aef8874) reveals that modern humans carry DNA from a previously unknown "ghost lineage" that interbred with our ancestors before the out-of-Africa expansion. The lineage contributed 0.5–1.1% of current human genomes, collectively covering nearly 1.5 billion bases—half the human genome.

Researchers at UC Berkeley developed a tool called TRACE to detect introgressed DNA segments. TRACE looks for sequences that are both old (in terms of mutation history) and young (in terms of recombination), a signature of recent introgression from a deep-diverging lineage.

How TRACE works

TRACE relies on ancestral recombination graphs (ARGs), which reconstruct the history of each base in the genome. The team used two ARG-generating tools and kept the one that produced better results. The combination achieved a false discovery rate below 0.25% and an accuracy over 90%.

To validate, they ran TRACE on African populations, which only received Neanderthal/Denisovan DNA via back-migration. TRACE found only 0.1% archaic ancestry in those genomes, consistent with the low error rate.

Key findings

• The ghost lineage split from modern humans over 800,000 years ago, roughly the same time as the Neanderthal/Denisovan split.
• Ghost DNA is present in all modern populations, but African genomes carry more diverse segments and some unique ones, reflecting the loss of diversity during the out-of-Africa expansion.
• About 100 genomic regions lack ghost DNA entirely in non-African populations.
• Ghost DNA appears in "deserts" that lack Neanderthal and Denisovan DNA, suggesting those regions are not universally intolerant to archaic variants.

Super-archaic introgression in Oceania

The team also found evidence of a "super archaic" lineage that split from modern humans nearly 1.8 million years ago, likely Homo erectus or a similar ancestor. In genomes from Oceania, about 0.3% of Denisovan DNA traces back to this super-archaic lineage. This is a tiny fraction of the genome, but it confirms a deep interbreeding event.

What it means for developers

For bioinformatics developers, this research demonstrates a practical approach to detecting hidden population structure in genomic data. The TRACE tool is a concrete example of using ARGs to identify introgressed segments with high specificity.

If you're working on population genomics, consider integrating ARG-based methods. The field is moving beyond simple reference-based mapping to full graph-based inference. Tools like TRACE (though not yet open-sourced) show the power of combining multiple ARG generators to reduce false positives.

Takeaway

This study settles a long-standing question: there is indeed a third archaic lineage in our genomes. The ghost lineage contributed more DNA than previously thought, and the super-archaic signal in Oceania adds to the complexity of human evolution.

For developers, the key is the methodology. Ancestral recombination graphs are becoming a standard tool for understanding evolutionary history. Expect more tools like TRACE to emerge, and consider how these techniques might apply to other species or even cancer genomics.