The Mirror in the X
There is a stretch of your genome that is almost entirely free of Neanderthal DNA. It has been known about for years. Geneticists call the gaps "Neanderthal deserts," and for the better part of a decade, the explanation seemed obvious: the Neanderthal variants that ended up on the X chromosome were probably harmful — genetically "toxic" — and natural selection scrubbed them away. A clean, structural explanation. Case closed.
It wasn't closed.
A paper published in Science on 26 February by Alexander Platt, Daniel Harris, and Sarah Tishkoff at the University of Pennsylvania has turned the story inside out. They looked at the question from the other side — not at Neanderthal DNA in modern humans, but at modern human DNA in Neanderthals. Three high-quality Neanderthal genomes exist: specimens from the Altai Mountains, Chagyrskaya Cave, and Vindija Cave, spanning 122,000 to 52,000 years ago. All carry traces of earlier contact with Homo sapiens, from an interbreeding event roughly 250,000 years ago.
On the autosomes — the twenty-two pairs of non-sex chromosomes — modern human DNA appeared in the expected scattered fashion. Then the team looked at the X chromosomes. And what they found was a mirror.
Where modern humans have a Neanderthal desert on the X, Neanderthals had a modern human surplus. Sixty-two per cent more human DNA on the X than on the autosomes. Not less, not the same — dramatically more.
The Moment the Hypothesis Snaps Into Focus
If genetic toxicity were the explanation — if Neanderthal X variants were simply harmful to humans and vice versa — then the pattern should be symmetrical. Both species should show deserts on their X chromosomes. But the pattern is anti-symmetrical. One species has a desert; the other has a flood.
There is really only one family of explanations that produces this shape: directional mating. The asymmetry of sex chromosomes makes it visible. Females carry two X chromosomes; males carry one X and one Y. A father can pass his X to daughters only — never to sons. When Neanderthal males mated with human females, fewer Neanderthal X chromosomes entered the human gene pool (the father's X goes only to daughters, while his Y goes to sons who carry no Neanderthal X at all). Conversely, more human X chromosomes entered the Neanderthal population, because the human mother always contributes an X.
The team ran simulations. Biased migration — even the extreme scenario of an entirely female human migrant population — could only produce a 1.3-fold excess. The observed excess was 1.6-fold. The numbers required something stronger: a persistent mate preference, operating not just during the initial interbreeding event but across subsequent generations. Males with Neanderthal heritage were more likely to find partners than females with Neanderthal heritage, within predominantly human populations.
The preference left its signature in the X chromosome for a quarter of a million years.
People, Not Populations
"Human geneticists have often taken a bizarrely clinical approach when looking at ancient genomes," Platt told reporters. "But these are all people, and we know that people have bias, and we know people have preferences."
The paper is deliberately agnostic about what drove the preference — attraction, social structure, competition, coercion. The debate among commentators has been less restrained. Steven Churchill at Duke suggested the pattern is hard to reconcile with anything other than hostile interaction between the groups. April Nowell at the University of Victoria pushed back: there is no archaeological evidence of interspecies violence at the relevant time periods, and in evolutionary biology, females are typically the choosier sex. It is at least as plausible that Homo sapiens women selected Neanderthal partners as that they were taken by force.
We do not know. The genome records outcomes, not motives. What we know is that the encounters were not random, that they were not symmetrical, and that whoever those people were — meeting at the edge of an ice sheet or in a river valley between territories — they had preferences that shaped the genetic architecture of every non-African person alive today.
The Genome as Social History
What strikes me most about this paper is the inversion it performs. For a decade, the Neanderthal desert on the X chromosome was read as a biological story: bad genes, purged by selection. A process indifferent to individuals. The mirror told us it was a social story: who approached whom, whose children were accepted, whose were not. The genome, read carefully enough, holds not just a record of what survived but something closer to a record of who chose whom.
This has been a recurring theme in the papers I have written about here. We assumed language was structured by innate grammar; it turned out to be shaped by cognitive ease. We assumed ancient humans lacked information systems; they had proto-writing forty thousand years ago. We assumed the brain stored language as formal trees; it also stores messy, frequency-worn footpaths. In each case, the elegant, structural explanation gave way to the messier, more human one.
The Platt paper extends the pattern into deep time. What we thought was the clean hand of natural selection acting on incompatible genes turns out to be the untidy reality of mate preference — of desire and avoidance, of social norms we will never recover, operating across species lines a quarter of a million years ago.
The X chromosome remembers who loved whom. Or at least, who chose whom. The distinction is as old as the question, and the genome cannot answer it. But it remembers.
Source: Platt, A., Harris, D.N. & Tishkoff, S.A. (2026). "Interbreeding between Neanderthals and modern humans was strongly sex biased." Science 391(6788). DOI: 10.1126/science.aea6774