Tens of thousands of years before written history, something consistent was happening at the edges of human migration. When Neanderthals and modern humans encountered each other as our ancestors expanded out of Africa, the mating that followed was not, it seems, random. A new study from the University of Pennsylvania suggests that Neanderthal males showed a strong and sustained preference for modern human females, and the evidence is written directly into ancient DNA.
The research, led by Alexander Platt, Daniel N. Harris, and Sarah Tishkoff, examined the X chromosomes of the small number of completed Neanderthal genomes currently available. Their findings, published in Science, reveal a striking excess of modern human sequences on those X chromosomes. That pattern mirrors what geneticists already knew about modern human genomes, which contain stretches with virtually no Neanderthal DNA at all. Scientists refer to these gaps as "Neanderthal deserts," and the largest of them spans the entire X chromosome.
The mechanics of inheritance on the X chromosome make it a particularly revealing record. Males carry only one copy of the X, inherited from their mothers, which means any disadvantageous gene on that chromosome has nowhere to hide. It faces immediate evolutionary pressure with no backup copy to compensate. This raises the first possible explanation for the Neanderthal desert on the X: straightforward natural selection weeding out incompatible Neanderthal genes.
Genetic incompatibility between two populations that have been evolving separately for hundreds of thousands of years is entirely plausible. Proteins interact in complex networks, and genes within those networks tend to co-evolve. Reintroduce an ancestral version of one gene from outside the network and it can disrupt the whole system, reducing the fitness of whoever carries it. The researchers did find some correspondence between Neanderthal deserts in modern human genomes and modern human deserts in Neanderthal genomes, which is consistent with this kind of mutual incompatibility.
But incompatibility alone does not fully account for what the researchers found on the X chromosome. The modern human sequences that appear there have a lower than average density of functionally important regions, such as gene-regulating sequences and protein-coding segments. That makes it harder to argue those sequences were preserved purely because they offered a selective advantage to Neanderthals who carried them. Less functional DNA being retained suggests something other than selection is driving the pattern.
That points toward the second explanation: biased mating. If most pairings between the two groups involved Neanderthal males and modern human females, fewer Neanderthal X chromosomes would circulate through the population. A male passes his X chromosome to only half his offspring, specifically his daughters. A systematic preference for modern human females, sustained across many generations, could gradually flood both populations with modern human X sequences.
The frequency of modern human DNA on the Neanderthal X is high enough that a simple one-directional preference is not quite sufficient to explain it. The researchers concluded that the offspring of these cross-species pairings must also have been preferred mates within the Neanderthal population over subsequent generations, compounding the effect. The authors are careful to note they did not rule out more complicated scenarios combining both selection and sex-biased mating, with natural selection potentially acting as a modifying force on top of the dominant sex-bias signal.
The study draws on published Neanderthal genomic data and compared the Neanderthal X chromosome patterns against X chromosomes from African populations, which carry very little Neanderthal DNA and therefore serve as a cleaner baseline. The contrast between the X chromosome results and the rest of the genome was sharp and consistent.
For those who follow the fast-moving field of palaeogenomics, the findings add another layer to an already complex picture. Research from institutions including the CSIRO and international collaborators has increasingly shown that ancient human history involved far more mixing between populations than the simple "out of Africa" narrative once suggested. Neanderthals, Denisovans, and anatomically modern humans were not neatly separated species on isolated evolutionary tracks. They met, they mated, and the consequences of those encounters are still present in living human genomes today.
What the University of Pennsylvania study cannot yet resolve is why this preference existed. Was it driven by competition for resources, by social structures within Neanderthal groups, or by something else entirely? The genomic record can tell us that something happened with remarkable consistency; it cannot yet tell us what motivated it at the level of individual behaviour across ancient populations.
The honest answer, as is so often the case in palaeogenomics, is that the evidence supports a strong hypothesis while leaving meaningful uncertainty intact. The X chromosome data is compelling. The interpretation of preferential mating is well-supported. But ancient human behaviour was almost certainly as complex and varied as modern human behaviour, and reducing it to a single pattern risks oversimplifying a story that played out across thousands of years and countless encounters. What the researchers have achieved is a rigorous narrowing of the possibilities, which is precisely how good science is supposed to work.