From Europe to Siberia, Neanderthal remains have gradually accumulated since the first 1856 find in Germany’s Neander Valley. Not only are there plenty of them, there are many near-complete skeletons. So we know a great deal about them, including their intimate relations with anatomically modern humans when the two met outside of Africa. Modern people throughout Eurasia are hybrid as a result. The Denisovans, on the other hand, were only discovered through genetic analysis in 2008 of a finger bone and a few teeth from the eponymous cave in western Siberia. Significant traces of AMH-Neanderthal interbreeding emerged in 2010 from segments of their DNA (between 1 to 4 %)  being found in living people whose ancestry lies beyond sub-Saharan Africa. Within a few months the same kind of genetic exploration revealed that AMH and Denisovans also interbred. It is now known that the DNA of indigenous people living on Pacific islands and in Australia, South, Central and East Asia and the Americas contains Denisovan segments (up to 5% in the Philippines)

Until 2019 the only tangible remains of Denisovans amounted to a finger bone, a toe bone and three molars. They were then joined by a partial lower jaw from a Tibetan cave, containing one tooth that yielded protein data that roughly matched those from Denisova Cave: not as convincing as DNA, but sufficient. About a decade ago Taiwanese fishers of the Penghu Channel between Taiwan and China began to collect ancient bones from their trawl nets, including those of hominins. In 2025 proteomic data from a Penghu hominin mandible again showed plausible Denisovan characteristics. For years Chinese palaeoanthropologists had been pondering on ancient, very robust skulls that had been discovered then hidden in the 1930s, and only recently re-emerged. One from Harbin, dubbed Homo longi or ‘Dragon Man’ had been dated at 146 ka. It had abundant dental calculus (plaque), which yielded mitochondrial DNA that matched that of Siberian Denisovan remains. At last the Denisovans had a face! There are many similar skulls in China, which had previously been assigned to H. erectus or a range of local ‘species’ but are yet to be subject to genetic and proteomic analysis. So it is possible that Denisovans may be ‘hiding’ in plain sight!

Now, once again thanks to Taiwanese trawling (and research by joint Japanese and Taiwanese consortia) two hominin upper-leg bones –a femur and a tibia – have entered the Denisovan collection from the seafloor off Penghu.  They are about 45 ka old, and assigned to two separate Denisovan individuals using the proteins of bone collagen. This is a particularly interesting development because upper leg bones are a guide to stature (Yousuke Kaifu and 11 others 2026. Denisovan leg bones from Taiwan reveal large body size. bioRxiv, preprint; DOI: 10.64898/2026.08.07.743438). The leg bones are the largest known from East Asian Pleistocene hominins: i.e. H. erectus and H. sapiens. The femur suggests an individual ~1.8 m tall, weighing around 83 kg, the tibia someone larger (1.9 m and 91 kg). Other commentators have suggested a stature of up to 2.3 m. A ridge of the femur known as its pilaster, to which major muscles are attached, is especially pronounced. This suggests enhanced mobility akin to that in modern hunter-gatherers. A separate preprint about the Penghu leg bones (Minoru Yoneda et al. 2026. Habitat and feeding ecology of a Denisovan from Late Pleistocene Taiwan. bioRxiv, preprint; DOI: 10.64898/2026.08.07.743455) uses carbon and nitrogen isotopes in the tibia to assess the individual’s diet. He/she ate a lot of terrestrial animal protein, but little in the way of aquatic resources. The late Pleistocene ages of the leg bones strongly suggests that the individuals may have encountered modern humans, and the two papers speculate on possible gene flow between Denisovans and AMH.

Let’s hope that this work encourages a flurry of research into other Pleistocene hominin bones from China. Proteomics are not as revealing as ancient DNA, but proteins are sufficiently robust to remain undecayed for hundreds of ka in climates that DNA does not survive for long

See also: Hawks, J. 2026. Legs of the last Denisovans. Personal substack, 8 August 2026.